Skip to main content

Full text of "The unity of Plato's thought"

See other formats






_ Cornell University Library 

B 395S55 U5 

Unify of Plato's fhoughf 

,. 3 1924 032 292 108 
°''" Overs 


I ^W^ 












The original of this book is in 
the Cornell University Library. 

There are no known copyright restrictions in 
the United States on the use of the text. 








The University of Chicago Press, Chicago }7 
The University of Toronto Press, Toronto 5, Canada 

Copyright 1903 by The University of Chicago 

New Impression 1960 

Printed in the United States of America 


Paul Shobet 


DuEiNG the past twenty years Platonic Forschung has come to mean the investi- 
gation of the relative dates of the dialogues by the statistical study of vocabulary and 
idiom. The general trend of modern philology and the reaction against mystical and 
metaphysical Platonism favored this tendency, and the work would perhaps not have 
been done at all if the workmen had not cherished illusions as to its value. To 
combat these illusions or to test in detail the logic of Sprachstatistik is not the purpose 
of this paper. A merely negative attitude toward any harmless form of human 
endeavor is unfruitful. But granted, since life is short, all that is claimed by the 
enumerators of KaOdirep and tC fi^v, the essential quality of Plato's thought remains for 
some Platonists^ a more interesting topic of discussion than the conjectural chronology 
of his writings. It has become the fashion to assert that the one depends upon the 
other, that we cannot interpret Plato's philosophy until we have determined the 
historic sequence of the dialogues, and with it the true order of development of his 
thought. But we have always known that the Laws and Timceus are late, that the 
Republic belongs to Plato's full maturity, and that the minor Socratic dialogues are as 
a whole presumably early. To affirm that more is necessary is to beg the question; 
it is to assume the very point in controversy that the philosophy set forth in the 
dialogues did develop in the sense required by the argument. The question is partly 
verbal. Every man's thought is developed out of nothing somewhere between infancy 
and maturity. Any author whose literary activity, like that of Plato, extends over 
half a century undergoes many minor changes of opinion, and reflects many varying 
moods of himself and his contemporaries. But it is not true of all, or of a majority, of 
the world's great thinkers that their first tentative gropings toward a philosophy and 
a criticism of life are depicted as in a votive tablet in their earliest published writings, 
or that the works of their riper years present a succession of shifting and dissolving 
views. Yet something like this is the assumption made by the increasing number of 
investigators who, in emulation of the triumphs of the statistical method, are endeav- 
oring to confirm, refute, or correct its results by a study of alleged inconsistencies, 
contradictions, or developments in Platonic doctrine. Abstractly the followers of this 
method would probably repudiate the principle here attributed to them. In their 
practice the desire for striking arguments and definite results leads them to assume 
that Plato was capable of producing a masterpiece like the Protagoras before his most 
characteristic philosophical and ethical conceptions had taken shape in his mind, and 

1 Notably for Bonitz ; see the judicious observations in Platonische Studien, 3d ed., pp. 270 B. and passim. 

The Unity of Plato's Thought 

that throughout the period of his maturest writings his leading ideas were in a state 
of Heraclitean flux, or were being casually developed from year to year. This method 
misleads scholars of great acumen and erudition to make false points, to labor fantastic 
analogies, and to cite irrelevant parallels. It betrays them into misplaced emphasis, 
disregard of the context, and positive mistranslation. In short, it necessitates the 
systematic violation of all the canons of the simple, sane, and natural interpretation of 
literature." Plato avoided rather than sought a rigid technical terminology, and 
prodigally varied the language and imagery in which he clothed hia most familiar 
thoughts. Every variation of phrase and imagery is pressed to yield significant 
contradictions or developments. The most far-reaching conclusions are drawn from 
the different shades of meaning attached to such words as "opinion," "dialectic," 
"philosophy," "sensation," "reminiscence," "participation," "presence," "com- 
munion," freely and untechnically employed by Plato to suit the theme and context.' 
The absence in any work of explicit insistence on a thought is supposed to prove the 
absence of the thought from Plato's mind at the time, and as a consequence, we are 
expected to believe in the most incredible combinations of maturity and naivete within 
the same writing. Or we are taught that Plato's development, like some Sophoclean 
sentences, proceeds in the order aba, and consisted in the acceptance, the rejection, 
and the re-acceptance of the same idea. The most reckless assertions are made that 
certain elementary thoughts appear for the first time in certain dialogues. The 
emphatic introduction of a term or idea is, according to the exigencies of the theory, 
now taken as proof that it is a novelty, and now explained away as a mere dramatic 
artifice. The rapid outline of an argument is alternately regarded, according to the 
requirements of the "chronology," as an anticipatory germ or a later r6sum6 of the 
fuller treatment found elsewhere. Fantastic conceits or bare possibilities as to Plato's 
literary motives and polemical intentions are treated as absolute psychological and 
historical certainties and made the basis of serious arguments.* 

May there not be some Trpwrov i/reuSos involved in a conception that thus betrays 
its advocates? It is of course a priori conceivable that Plato's thought did unfold 
itself in this tentative and fumbling fashion. Examples of such mutations and nuta- 
tions can be found among the Pichtes and Schellings of modern philosophy. They 
are still more frequent, as Professor Gildersleeve has wittily shown, in the history 
of modern philology, and, as I may add, in the in;terpretation of Plato. But it is at least 
equally probable that Plato's philosophy and his conception of life had taken shape 
at the age of thirty or thirty -five, and that his extant works, though not of course a pre- 
determined systematic exposition, are the naturally varied reflection of a homogene- 
ous body of opinion, and of a consistent attitude in the interpretation and criticism of 

2 Examples throughout the paper. alized statements and criticisms o£ tendencies in the 

,., jT„„. ,,/-.■■ J ^ tu ^ thought of the time, and especially the hypothesis that he 

s Infra, and Lutoslatvski, Origin and Growth of .. • j ^ ■ ^ \^ . ,• 

. , _ . . satirized contemporaries under the names or earlier 

' Sophists. Such hypotheses will be wholly disregarded in 

* To this category belong nearly all conjectures as to the following study, as a mere hindrance to the apprehen- 

the particular philosophers referred to in Plato's gener- sion of Plato's own meanings. 

Paul Shoeey 

contemporary life. And if this were the fact, it would be a far more important fact for 
the interpretation of his writings than the determination of the relative dates of the 
Phcedo and Symposium or even than the demonstration that the Sophist, Statesman 
and Philebus follow rather than precede the Republic. I am not arguing against 
such a dating of the dialectical dialogues. I do not deny the value of the more vivid 
conception that we gain of Plato's later mood and manner by combining and compar- 
ing the traits of these dialogues with those of the Laws and Timceus. This is no 
a/37os \o'7o? directed against all sober critical investigation of the difficult problem of 
Plato's chronology. But the attempt to base such a chronology on the variations and 
developments of Plato's doctrine has led to an exaggeration of Plato's inconstancy that 
violates all sound principles of literary interpretation and is fatal to all genuine intelli- 
gence of his meaning. The implicit canon of this method is that variation in literary 
machinery and expression must be assumed to imply divergence or contradiction in 
thought. To this I wish to oppose an interpretation based on the opposite canon: 
that we are to assume contradiction or serious alteration in Plato's thought only in 
default of a rational literary or psychological explanation of the variation in the form 
of its expression. As Professor Maguire says in his forgotten but very acute essays 
on the Platonic ethics : " If we are anxious to find out inconsistencies in appearance, 
we shall find them in abundance. But the student of Plato will perhaps discover that 
it is more fruitful, because more philosophical to commence with the points of agree- 
ment." The ultimate test of the two methods must lie in the appeal to specific texts 
and contexts, and there will be no lack of this in the following pages. But by way of 
preparation it is first advisable to enumerate some of the general features of Plato's 
writings that make the sane and simple literary interpretation of his meaning so diffi- 
cult and so rare. 

1. Plato is not only a thinker, but also a dramatic artist and an impassioned moral 
and religious teacher. Although, as Schopenhauer says, he is really the most severe 
and consistent of logicians, and holds the threads of his design in an iron hand, his 
dramatis personae affect to follow whither the argument blows,* and he often seems 
more concerned to edify or entertain than to demonstrate and conclude. Wherever 
his aesthetic or moral preferences are involved he cavils on terminology and breaks 
into seemingly irrelevant eloquent digressions in a Ruskinian fashion sorely puzzling 
to those not in sympathy with his mood. If forced to accept the substance of a repug- 
nant theory, he translates it into language more consonant with his feelings. This 
peculiar mixture of rhetoric and logic, of edification and science, misleads both the 
sentimentalist and the scientific puritan. The one often mistakes the ornament for 
the substance, the other distrusts perfectly sound reasoning because of his distaste 
for its emotional accompaniment. 

Again, Plato stimulates our own speculation in so many ways that we are apt to mis- 
take the drift of his meanings not because it is not clearly defined, but because we abandon 

s Not only in the earlier dialogues, but in Bep., 394 D ; Thecetet., 172 D ; Laws, 667 A. 

6 The Unity op Plato's Thought 

it to pursue our own. The clever essayist tells us what he himself thought db propos of 
this or that brilliant suggestion. The investigator too often begins by selecting a few 
detached notions and formulas as adequately representative of each dialogue, and then 
proceeds to juggle with ingenious combinations of these and .the interpretations put 
upon them by his predecessors. Neither interprets Plato's real thoughts as they lie 
open to any competent reader who will patiently study him to the end and report the 
things on which he lays most stress." 

2. In the second place, Plato's dramatic quality affects not only the artistic setting 
and the personages, but the ideas which he brings upon the stage. Plato's serious 
meaning detaches itself with perfect distinctness for the faithful student. But the 
hasty reader is more likely than not to receive as Platonic ideas that have a purely 
dramatic significance ; or that are falsified by isolation from their context.' And the 
investigator in pursuit of a thesis too often attributes specifically to Protagoras, 
Antisthenes, EucKd, or Isocrates ideas that Plato has generalized and decked out 
beyond all recognition, as representatives of the spirit of the age. 

Again, arguing for victory, the maintenance of a thesis in jest to test an oppo- 
nent's metal or display one's own ingenuity was a common practice in the world which 
Plato depicts, and is frequently illustrated in his writings. The Platonic Socrates, 
under cover of an ironical profession of ignorance, employs a similar method to 
expose showy pretenders to universal knowledge, to produce a salutary conviction of 
ignorance, or to stimulate youthful thought, and prepare the way for a more serious 
analysis by an exposition of the antinomies latent in conventional opinions. It fol- 
lows that the ostensible failure to conclude an argument, the avowal of bewilderment 
and perplexity, the admission even of positive fallacies of logic in any given dialogue 
prove nothing as to the stage of development of Plato's own thought at the time. The 
hypothesis that the fallacy was intentional, and that the airopCa was affected for a 
purpose, has at least an equal claim to be tested by all the probabilities in each case. 

3. Expositors of Plato seem strangely oblivious of the limits thus far set to all 
systems of philosophy. They treat as peculiar defects of Plato the inconsistencies 
which they detect in his ultimate metaphysics after they have elaborated it into a 
rigid system which he with sound instinct evaded by poetry and myth. They 
habitually write as if they themselves and their intelligent readers were in possession 
of a final philosophy which reconciles all conflicting claims of metaphysical analysis 
and common sense, and from the heights of which they may study merely as a his- 
torical phenomenon Plato's primitive fumbling with such problems as the nature of 

6 Such a reader is Bonitz for the most part in his ad- were intended seriously, and not a few continue to quote 

mirable analyses. The<Btet., 156 S., as Platonic doctrine. Under this head 

'A notable example is Herbert Spencer's inference fall most of the " fallacies " discovered in Plato: those of 

from Bep., 339 D, that Plato, like Hobbes, makes state *^^ Parmenides, which, as we shall see, are intentional; 

enactments the source of right. So President Eliot has t^^ose of the Gorgiaa, dramatically justifiable against the 

been recently misled by Zelleb's misuse of Bep., 421 A extreme thesis maintained by CaUicles; those of Bep., I, 

{Phil, der Griechen, 4th ed.. Vol. H, No. 1, p. 890), to prove 333 E, and 349 B, which Zeller (p. 652) thinks Plato did not 

that Plato would not educate the masses. Many scholars perceive, 
still seem, to think that the etymologies of the Cratylus 

Paul Shobey 

universals, the antinomy of unity and plurality in thought and things,' the relation of 
mind and body, the possibility of a consciousness of self or a knowledge of knowl- 
edge, the proof of immortality, the freedom of the will, the difficulty of conceiving or 
defining good except in relation to evil, the alternative of excepting thoroughgoing 
relativism and phenomenalism or of positiag a noumenon that cannot be described or 
brought into intelligible relation with phenomena. We are told that he has "keine 
Ableitung des Sinnlichen," as if there were somewhere extant a satisfactory deduction 
of the sensible world from some higher metaphysical principle. It is objected that 
the relation of the ideas to the Deity is undefined, and that the personality of God is 
not investigated, as if any results could follow from an attempt to define the relation 
of the metaphysical noumenon to the Deity, or from an investigation of the person- 
ality of God. The absence of a complete table of categories is taken as a defect in 
Plato's system or as a proof of the immaturity of the Phcedrus, as if the Aristotelian 
and Kantian categories were not mere illusions of the metaphysical instinct, and Plato 
was not far wiser in proposing only such categories and classifications as the argument 
in hand required. 

A chief merit of Plato is that he clearly recognizes and sharply defines the limits 
of scientific thought in these matters. When the interests of the^ moral and religious 
life, as he conceives them, are at stake he resorts to myth to express his hopes and 
aspirations. Where the epistemological problem compromises the foundations of prac- 
tical certainty and sound method, he arbitrarily postulates the solution that will best 
serve his chief purpose — the extrication of a practicable working logic from the hope- 
less dialectical muddle of his time. But he is always careful to distinguish his neces- 
sary practical postulates from his mythical and metaphysical assumptions.' The 
dogmatism of his later works has been as much exaggerated as the Socratic doubt of 
the minor dialogues.'" 

4. As a fourth cause of misapprehension we may count a certain quaint and curious 
subtlety in the use of abstraction and antithesis characteristic of all Greek writers, but 
carried to its farthest extreme in Plato. His reasoning often proceeds by what 
seem to us excessively minute verbal links. This is generally thought to mean 
merely that the modem mind has learned to abridge the formal process by taking some 
things for granted. But it is often due to Plato's anxiety to anticipate the cavils and 
quibbles of the age before logic ; or his wish to bring out neglected shades of meaning. 

Again, Plato, like all serious reasoners, employs unreal abstractions to express 
ideals and test hypotheses by extreme cases." But in addition to this the Platonic 
Socrates meets a fallacious and fantastic abstraction from the conditions of reality, not 

'Astonishment is often expressed at the attention apodictic replies in the "later" works proves nothing that 

bestowed by Plato upon the problem of the one and the is not already involved in the fact that they are not dra- 

many, as if, transferred to psychology, it were not still matic disputations. A consenting respondent naturally 

the crux of all our metaphysics. gives "apodictic " answers. 

'Jfeno, 86 B; PAccdr., 252 C, 265 C, 274 C ; B^., 416 BC, "E.g., the isolation of pleasure and intelligence in 

517 B, 506 C. Phileb., 21, to which Grote objects. 

10 Tim., 72 D, Laws, 641D, 799 D, 812 A. The percentage of 

The Unity of Plato's Thought 

by exposing the fallacy, but by translating all the real facts into the language of 
abstraction. There is no real fallacy in such procedure, but a sense of fallacy results 
for the modern reader.'^ Allied to this is the use or abuse of antithesis. Opposite 
views are first stated with ruthless consistency in their most abstract and extreme 
form. And the truth is approached through a series of compromises and mediations.'^ 
Dramatically, Plato is right. This is the course of discussion among ordinary men in 
all ages. But the elaborate refutations which Plato thinks fit to give of the crudest 
form of hostile theories sometimes produces an impression of unfairness upon modern 
critics." They forget two things : first, that he always goes on to restate the theory 
and refute its fair meaning ; second, that in the case of many doctrines combated by 
Plato there is no evidence that they ever were formulated with the proper logical quali- 
fications except by himself.'^ 

5. In the fifth place, and finally, we may mention the difficulty of confining 
the infinite variety and suggestiveness of Plato's thoughts in the framework of 
any system either of philosophy or of exposition. It is possible to present 
Plato's ethical and social ideals in a fairly systematic r6sum6. The theory of 
ideas may be restated in the Platonic terminology, which does not teach us much, 
or analyzed in relation to the underlying psychological and ontological problems. 
Special chapters might be written on Plato's attitude toward inchoate physical science, 
the temper in which he faced the religious problems of an age of transition, his portrayal 
and criticism of the literary and artistic life of his time. But a complete system 
of philosophy with principles subordinate, derivative, and interdependent, and a fixed 
technical terminology, cannot be extracted from the Platonic writings. This will not 
greatly grieve those who are aware of the perfect futility of all such system-building, 
even when the architect possesses the genius of a Spinoza, a Kant, or a Schopen- 
hauer. But the expositor of Plato can hardly avoid attempting to cast his exposition 
into some systematic form, and the recalcitrance of his material is to him a serious 
problem. No method is quite satisfactory. The atomism of Grote, Jowett, Bonitz, 
and Horn, that treats each dialogue as an isolated unit, is the renunciation of all 
method. The clever attempts of a succession of French expositors to deduce all Platon- 
ism symmetrically from a few principles are more ingenious than convincing." The 
exhaustive schematism of Zeller, applied alike to all philosophers from Thales to 
Plotinus, is philologically a masterly achievement of German erudition. But, though 

12 £. ff., in iJep., I, 346, the separation o( (iktSoitik^, the 451 E, 453B, 489 D, Similar is the treatment of Homo Jlfen- 

wage-earning power, from the other functions of each art sura in the Protagoras, and the claim of pleasure to be the 

and craft. chief good in the Philebus. 

^^Philebus, Thecetetus, Bep.,1 andll, Gorgias. 15 piato may have found hints and suggestions of the 

HE. g., in. the Cratylus, S&i A, thetheory that language views he brings on the stage in Euripides and the Sophists 

is a mere convention is first stated in the most extreme (DOmmlee, Prolegomena zu Platans Staat). But so far as 

form. In the Gorgias a long argument is spent to drive '^® know, he is the first thinker who could present a com- 

Callicles from a position which he affirms was assumed in Ple'e logical statement of any philosophidal theory in all 

jest (499B). In iJep., 338 C, Thrasymachus's definition of its bearings. 

justice is taken in a grotesquely unfair sense in order to 16 See my review of Hal^vy, ThAorle platrndcienne des 

force him to state it more clearly. Cf. Laws, 714 C; Gorg., sciences, Philosophical Review, Vol. V, p. 522. 

Paul Shorey 9 

rarely admitting gross and palpable errors, Zeller's exposition frequently misses the 
true proportions, perspective, and emphasis that would be brought out by a more 
flexible literary and philosophic interpretation. 

The present study, though it touches on most topics of the Platonic philosophy, 
does not attempt a complete historical survey. Some subjects I have discussed else- 
where. There are many details (in the Laws and Timceus, e. g.) which would be 
irrelevant to the main purpose of emphasizing the unity of Plato's thought. The order 
of presentation adopted after many attempts is a compromise between the systematic 
and the atomistic. The Platonic ethics, the theory of ideas, and an outline of the 
psychology will first be set forth as a whole. A group of logical and metaphysical 
problems will be discussed in connection with the Sophist and Parmenides. Other 
topics and some repetitions from a different point of view will follow in a survey of the 
principal dialogues taken one by one. 


The chief topics of the Platonic ethics are these: (1) the Socratic paradoxes; 
(2) the definition of the virtues, and, more particularly, the determination of their 
relation to a postulated supreme science or art, to happiness, to the political or royal 
art, to the idea of good; (3) the problem of hedonism; and (4), associated with it, 
the attempt to demonstrate the inseparability of virtue and happiness." 

1. Plato always formally maintained that all wrongdoing is involuntary;" that 
virtue is insight or knowledge, is in its essence one, and can in some sense be taught." 
Sometimes he merely dramatically illustrates the conflicts that arise between these 
paradoxes and common-sense. Elsewhere, most explicitly in the Laws'" but by impli- 
cation even in the minor dialogues, he reveals his perception that these propositions 
can be reconciled with experience only by the conscious employment of words in a 
special sense.^' Wrongdoing is involuntary (1) because all men will the good or what 
they deem the good;''^ (2) because no man who knows the right will do the wrong, if 
we take knowledge in the highest sense, or refuse the term to any cognition that does 
not control the will;''' (3) because the conditions that shape conduct lie far more in 
heredity, education, and environment than in our conscious wills." The contradiction 
noted by Aristotle between this charitable principle and the edifying proclamation 
" virtue is free," " is emotional rather than scientific.^^ The modern free-will contro- 
versy arises out of two conceptions not connected with this problem by Plato: the 

" These are, as a matter of tact, the chief topics of the 20 689 D, 696 C, 710 A, ijv nt atiivvvuii' iv Xeyoi, i^povTivtv 

ethical dialogrues. If we base Plato's ethics on the idea of irpoiravayKdiitiv elvau to aia^povtlv. 

good, or on any other metaphysical principle or schema- 2iiacAe«, 196 Ej iocftes, 191E, kvipeloi. . . . . ev ^Sovais, 

tism, we shall distort his meanings. cf. Laws, 633 D E, and Sep., 429D; Rep., 443E, «4A; 

isXen., Mem., 3, 9, 4; 4, 6, 6; Apol., 26A; Protag., 345 D, Themtet., 176 C; Polit., 306A. 
358CD; Meno, 77,78; Gorg., 466E, 467B=iJep.,577E=Z,aM)s, nueno, 77; Euthydem., 279; Symp., 205A; Gorg., 468. 

688B;iJep 382A(?), 413Aa),492E(1),589C; Pft«e6,22B; .sp.otoff., 352B; ia™, 689; The<etet.,mC. 

SopA., 228C, 230A; Km., 86D; ioMJS, 731C, 734B, 860D. w, , 

,„ „ ^, J nonr, r »ii A ■ •■ • a- s 2*rjnj., 86D. 25 iJep., 617 E, ipeTJl 5« iSeo-irOTOl'. 

WEuthydem., 2&2C; Laws,6iiA, us 01 yt opewirciTaiSev- i...-., w^. x >,j,., « . , r 

/Ac^ot ffxeSoj' aya^ol ■yt-yfoi'Tai. 26 Cf. my note in A. J. p.. Vol. X, p. 77. 

10 The Unity of Plato's Thought 

infinite foreknowledge of God, and the absolute continuity of physical causation. It 
is, then, unprofitable to inquire whether Plato taught free-will or determinism."' 
But it should be distinctly noted that in the Laws he employs precisely the logic of 
modern determinism to prove that the involuntary character of wrongdoing is com- 
patible with the distinction for legal purposes of voluntary and involuntary acts.^' 

Virtue is knowledge because it must be assumed to be a good, and the only certain 
good, the only sure guide to the good use of what the world calls good, is knowledge.^" 
Opinion and habit may often suffice to regulate action, but persistent right opinion 
presupposes knowledge in its teachers, and the highest rule of conduct must be 
deduced from and referred to a rational apprehension of ultimate good.'" Virtue is 
one because each of the virtues is a form of knowledge,'" or because each, when taken 
in the highest sense, involves all the others.'^ Virtue is teachable in the senses in 
which knowledge and right opinion may be taught. The capacity for knowledge, the 
divine faculty, is innate, but teaching and guidance may direct it toward the good.'" 
The ordinary virtues of habit and opinion may fairly be said to be taught when they 
are systematically inculcated by superior wisdom enlisting all the forces of society ^n 
its service."* This is not the case at Athens,'^ and therefore the Platonic Socrates 
alternately affirms and denies the possibility of teaching "virtue,""* and at the close 
of the Meno declares that under present conditions it comes by a grace divine which 
is equivalent to chance."' 

Plato uses, but is not himself confused by, the Socratic analogy between the 
virtues and the arts and sciences."" That comparison, though it ignores the distinctively 
ethical element, contains a certain measure of truth. In a sense, each of us is good 
in that which he knows. "^ Knowledge as ordinarily understood is not virtue, but it 

27ZELLEE, p. 853; JowETT, Vol. Ill, PP. 408, 425. to knowledge, opinion is imparted iy rH iraiSeU, 429C, i. c, 

28 861-864 C. The meaning of the passage, though often '^ virtually taught, 

misunderstood, is perfectly clear, and Plato warns us, 34j;ep., 500D, 429CD; Poirt.,'309 D ; Laws, passim. 

864B, not to catil about the terminology. 35jJep., 492E; Tim., 87B; Meno, 93Bffl.; Protag., 320; 

2«Ev.thydem., 281, 289; Merw, 88C. Cf. from another ^«i'- 520B -Euthyphro, 2CD; Gorg., 521D; Apol, 24,25; 

point of view Fhoedo, 69 AB ; Protag., 356, 357, with Fhileb., Laches, 179 C D. 

41E. ^Protag.; Meno; Buthyd.,2&2G {274E). 

sojlfeno, 97B; Meno, 100 A, olos koI aK\ov n-oi^crai, eti-. Cf. 37 For this interpretation of «ecV (loip^i see Magciee, p. 

Euthyd., 292D; infra, p. 16: Laws, 951 B. 63, andZELLEE's full refutation of other views, p. 594, n. 4, 

T^ pp . Rep., 492, 493. At present good men spring up avTo^arot 

nLaches; Protag.; Phmdo, 69AB. Meno, 71 D ff., is (jj^^ _ gjOB; cf. Protag., 320A; Euthyd., 282C); even in 

logical rather than ethical. The unity of aper, is postu- ^j^j^^^ gj^jg^^ ^^^^^ gj^ ^ ^^^ ^^.^. ^^^^^ „. ^„^^^, 

lated, like that of any other abstract idea, as a precondi- ^„^^„„, ^^jj^ ^sj.\ov iv ^iv^i^ovixivai, ^6K^„,v ij «ai ^r,. 

tion of a definition. „„mi » -r^. . . , . ^. 

38 The lesser Htppias (certainly by Plato) presents the 

31 Gorg., 507 A; Laws, 696 C. There is a suggestion of fallacy inits most paradoxical form (the voluntary lie better 

this also in the (of course intentional, BoNlTZ, Platonic ji^an the involuntary) and by its obvious irony (372 DE, 

Sttidies, p. 265) faUacies of Protag., 330, 331. 376 C) shows that Plato " already " in the Socratic period 

33 iJep., 518B, 519A. This apparently contradicts the does not take it seriously, but merely uses it for dramatic 

statement of the Meno, 99A, and.Protojr., 361B, that ini- or propadeutio purposes. Zelleb, p. 597, takes this as 

<rT^/ii) alone can be taught. But the objection is captious. Plato's real opinion, citing Sep., 535 D and 382, which 

The Republic is satirizing the exaggerated claims of the merely use the paradoxical terminology to emphasize the 

Sophists and is speaking of the faculty, not the content, of thought, acceptable to Mill or Huxley, that the mere intel- 

knowledge. The whole higher education is a teaching of leotual love of truth (knowledge) ought to be counted a 

knowledge in a sense. And, on the other hand, though virtue as well as the ordinary virtue of truthfulness, 

both Plato and Aristotle limit teaching in the strict sense 39i(ic7se«, 194D; Lysis, UQT) ; Sep., 349 E. 

Paul Shoeet 11 

does away with many forms of wrongdoing. It is not courage, but the man who knows 
how is less likely to be afraid.*" It is not crax^poavvr), but it is incompatible with many 
forms of a^poa-vvTj. The wise man knows his own limits, and will undertake only what 
he can perform." Partly for these reasons, and partly because he did not or, in ironical 
assumption that others were even as himself, would not recognize that men know the 
right and yet the wrong pursue, the Platonic Socrates seems to ignore the chief ethical 
factor, a virtuous will, and argues that he who knows justice is just." But such "fal- 
lacies" are for Plato merely the starting-point of a fuller analysis. All knowledge is 
good and commendable," but the supreme knowledge that may be identified with 
"virtue" is plainly something different from the specialties of the arts and sciences." 
Courage, for example, apart from mere animal and temperamental fearlessness, may be 
defined as knowledge of what is and is not to be feared. But this involves real knowl- 
edge of good and evil, a complete ideal of life, either that of the Sophists and average 
Athenian opinion, or that unfolded by Plato himself in the Republic. The attempt to 
define courage in the absence of these distinctions merely illustrates the inadequacy 
of conventional ethical thought." 

The effective application to these problems of the obvious distinction between 
science and right opinion requires the larger canvas of the Republic. And even then 
it remains true that the courage most worthy of the name implies a complete philo- 
sophic mastery of the conception of life that educates the masses in such right opinion." 
Plato tacitly assumes that this supreme knowledge will be inseparable from the vir- 
tuous will in his philosophic statesmen as it is in Socrates.*' And thus on this higher 
plane the Socratic paradox becomes true again. 

It matters little to the consistency and unity of Plato's thought whether we 
regard this harmony of the intellect and the will as a mere ideal or as a practicable 
postulate realized in Socrates and to be fulfilled by others in a reformed society. The 
distinction once drawn, the ideal once affirmed, Plato can afford to make concessions 
to common-sense. He can admit that in present experience a kind of bravery is 

w Laches, 193 ; Protag., 350. between the desires and the ethical convictions the grossest 

"Xen., Mem., 2, 2, 24; Charm., 171 DE; Ale., I, IITDE; torn of "ignorance." 
Sophist, 229 C ; Laws, 732 A. " Protag., 318 B ; Laches, 182 D. 

«Gors., 460 B. The fallacy, so far as it is one, is in- ** Charm., 165 G; Euthydem.,?&ZE,290; Protag., 311, 312, 

tentional. Observe Kara toOtoi' rbv AdYPi/, and the explana- 319 A ; Laws, 961 E ff. 

tion in Bep., 438 D E, that the knowledge of health, though 45 x^jg^es ; Protag., 349, 350, 360 D ; Bep., 429, 430. 

difierentiated from knowledge in general, is not neces- j o j • jonr. • i 1 - r,^ 

sarily healthful. C/. also the recognition of common-sense ^ The courage defined in 429C xs o^^ -^'"«Y V'- f^- 

in m-D, TO ^J. iUa.a uparrc. «.««.oo-v^v i^^o.^l. But for «W-S, ye,Laws,mA- PolU., 309E; Pfe<^o, 82A. There 

, ', , r, 4. J n,„ ,^„„„,-„, Ji. i„ are, strictly speaking, three or four grades ; brute animal 

the broad purposes of the argument of the Gorgias it is """ =-'''-^ J "»"' s' = . 

rue (460E)th^ rhetoric, if really the science of the just, courage the courage of soldiers and citizens in ordinary 

could not be the instrument of injustice which Gorgias states, the citizen courage of the Platonic state, the philo- 

with unconscious immorality complacently represented it sophic courage. 

to be. Socrates is 0I05 tUv enii/ iiriSeA aWa ireieeaiai. Ti r<f « This harmony is the chief point in the selections and 

\iyif,'crito, 46B; c/. Laches, 188 DE; Gorff.," 488 A. Hence, tests applied to them; Bep., 485, 486, 539 Dff. Cf. Polit., 

as Aristotle (Eth. nic., 7, 2, 1), quoting Protag., 352 B, says, 309 A B. The Laws emphasize character, as compared with 

he thought it monstrous that any other impulse in man inteUect, still more, and preserve the identity of the moral 

should prevail over his better knowledge. And Plato in and the intellectual " which are ever dividing, but must 

his latest work refuses the term " knowledge " to any belief ever be reunited " ( Jowett), by reserving the word " wise " 

that does not control the will, and pronounces discord for the virtuous, 689 D. 

12 The Unity of Plato's Thought 

found dissociated from the other virtues." He can allow the word aa^poavvri to be 
used merely for the instinctive temperamental moderation in appetite that is the for- 
tunate endowment of some children and animals/' He can recognize that knowledge, 
or at least quickness and acumen of thought, is not infrequently associated with 
intemperance and injustice.^ But he prefers to translate the facts into a more edify- 
ing terminology. Conventional virtue is a worthless currency unless redeemable and 
redeemed by and in the coin of wisdom." And, on the other hand, we will refuse the 
name of wise to him whose will does not follow his judgment of right; and we will 
grant it to the man who knows enough to obey his acquired belief in the good rather 
than the innate promptings of appetite, though he know not how to swim or recite 
the alphabet.^'' 

2. Plato found the suggestion of the cardinal virtues and of the predominance of 
justice in the poets. He also mentions oo-toT???" and fjxyaXoirpeTreia, the latter some- 
times with irony." But the number four was consecrated by its incorporation in the 
scheme of the Bepublic. This implies no change of doctrine. Even in the Republic 
other virtues are mentioned.^ And in the Euthyphro it is hinted that piety is a form 
of justice.^ 

Plato would always recognize piety as one of the chief virtues, or perhaps as a 
synonym of all virtue," and he would always shrink from giving so problematical a 
concept a place in a scientific scheme.'*' 

Several of the minor dialogues turn on the attempt to define the virtues and allied 
notions. The Laches and Charmides are both Socratic quests for definition — of 
courage in the one case, of temperance in the other. Both involve the antithesis of 
the quiet and the energetic temperament.^' Both terminate in perplexity — in the 
puzzle that, if any one virtue is identified with the supreme knowledge that will make 

48 Protagoras maintains this view, Protag., 350, and is of piety, I should accept that of Bonitz as formulated by 

not answered by Socrates, who refutes him only indirectly Peofessob Heidel (introduction to his edition of Euthy- 

by the proof that all virtue is one — the science of measur- phro, p. 24). It is the endeavor to realize the good felt as 

ing pleasure and pain. But the obvious fact of experience the service of God, and as a willed co-operation with Him. 

is presumably as clear to Plato when he allows Protagoras But this is a mood in relation to, or an emotional synonym 

to state it as when it is! enunciated more explicitly in the of, all virtue. It is not one aspect of virtue which it is 

PoiiticMS, 306 B, or the iaws, 631 C Zellee (p.599) incom- necessary to distinguish in relation to a special field of 

prehensibly affirms that the plurality in unity of virtue is conduct or a particular classification of the faculties of 

foimdonlyin the i?ep«6Kc,' the sold. 

49Z,a«)s, 710AB. ssThe suggestion that the Euthyphro "eliminates" 

60Be„ 519 A; Laws, 689 D, Bt. .rpo, t«xo! t^j ,(,vx?!; piety, and that the ilfeno may be dated by its recognition 

Themtet 176 C of oo-iiiTijt {78 D) is utterly fantastic. 

siPfccedo 69 B 59' C/. Cftarm., 159 B fE., with PoHt, 307 A B. Tempera- 

'' ' , , - „. m. J J ment is not virtue, but is the basis of the seeming opposi- 

M Laws, 689 D, Ml" VP«»'A'«Ta ^,^e -e.v. Cf. Themtet., ^.^^ between bravery and temperance {PoUt., 306, 307 ; Sep., 

176C,T<i;oJl..iS.KoB.r. na'P-? apttrr ex^t to m, <rvvx"pe'^ 410DE, 503CD; Laws, 735A, 681B, nSB, 831Ei Protag., 

5«r,5 irrb ^ovovpycw eirat. The whole passage IS in the ^^^^_ j^.^.^^ ^^^ Laches, for want of this distinction, 

mood and temper of the Laws. maintain opposite paradoxes. Socrates calls our attention 

53 Protag., 329 C ; Meno, 78 D ; Laches, 199 D. to this by attributing to Nicias the doctrine oiioim Uovra. 

54 MetM, 74 A ; Bei). , 560 E. In Meno, 88 A, e Vi«"« and ««' <'^<'*<>'' • ■ • • "^P" i-vip^iiiv - 7re*u«eVai, (196 E) . In the Be- 
V ■ are included public (430 B), Plato chooses to deny the term bravery 

'""""' ■ to mere animal courage. In the iajos, 963 E, he attributes 

^5 402 0, €A£v9.pidT>)!, neyaAoirpe>r«i,a536A. ^ ^j^^^j ^^ courage to children and animals. But oy^oim 

56 Cf. also Protag., 331 A. Trtijiuicei/ai pointedly ignores the distinction of tempera- 

oi If it were desirable to produce a Platonic definition ment. 

Paul Shoeet 13 

us happy, the distinction between the virtues vanishes;*" or in the tautology that the 
knowledge that is good is knowledge of the good.*' 

It is often assumed (1) that Plato was serious in these attempts to express by a 
phrase or a substituted synonym the essence of a virtue and the various and contradic- 
tory meanings of its conventional name ; (2) that the failure and pretended perplexity of 
Socrates at the close mark the point reached by Plato's own thought at the time. This 
is a priori conceivable. But the following considerations make it highly improbable : 

a) Plato, in this unlike Xenophon,*^ always proceeds as if he were aware of the 
true theory and use of the definition and of the multiple meanings of ethical terms. 
All attempts in his writings to work out abselute and isolated definitions fail.*' His 
own definitions, when not mere illustrations,** are always working hypotheses*' or 
epigrammatic formulas, subordinate to and interpreted by the argument of which 
they form a part, and recognized as imperfect, but sufficient for the purpose in hand.** 
The definitions of the virtues in Rep., 429 ff. cannot be understood apart from their 
context, and are never used again. They are declared to be a mere sketch — 
vTToypa^ijv, 504 D.*' How shall we explain this on the supposition that he was under 
any illusion as to the value of absolute and isolated definitions? 

b) Plato repeatedly refers in a superior way to eristic, voluntary and involuntary,** 
and more particularly to the confusion, tautology, and logomachy into which the vulgar 
fall when they attempt to discuss abstract and ethical problems.*' Some of these 
allusions touch on the very perplexities and fallacies exemplified in the minor 
dialogues.'* They do not imply that Plato himself had ever been so confused." Why 
should we assume that he deceives us in order to disguise his changes of opinion, or 

60 Laches, 199 E. dialogues cures. Cf. Meno, 84 A B. So Soph., 232 A B, gives 

61 CTiarm., 174 B; c/. Kep., 505BC — a connection gener- the raison d'etre of passages (.Gorgicn, Protag., Jon) in 
ally missed. which a pretender to universal knowledge is pressed for a 

62 The Xenophontic Socrates perceives no difficulties, specific definition of his function which he naturally is 
is never in doubt, and propounds dogmatically such defini- unable to give. 

tioas as y6^ilJ.oy = SiKaLov, Mem., iy,i, 12. ■'oPolit., 306 ff., especially 306 A, rh yap apcrJ,, ^epos 

63 Except the not quite serious definitions reached by »""?' "«" S;«*opoj- eW r.^a Tpd^ro.. To«jrepi Adyot, a^*.<.p,- 
dichotomy in the Sophist and PolUicus. Cf. Charmide,, :\"'^ «? ""^ '"/-"S^tov ^po, ra, ™. woKXu,. Jof«. Cf. Laws, 
Laches, Lysis, Meno, Theoetetus, Euthyphro, Hippias Major. «" »• '"^^''''"''r'" " ' ; " Z""" '■""'' '"' Z 

6*Taxo,, Laches, 192B; ..x^a, Meno, 75, 76; .,Ad,. Be^b., 3iS^, .^xo^..' a. re X.y..y .ara ra .o^.(.^^^^ 

/ ^ - -i • J ono T^ With reference to the arguments of Gor-ff., 474 C ff. Cf.Laws, 

Theoetet., 147 C ; .,Ato!, ibtd., 208 D. g^, ^^ ^.^j^ reference to the problem of the Lysis ; Laws, 

65PAoedr.,237D, 6MoAoyi?«e>ievoiopor. C/. 263DE. gg^ g^ gg^^ ggg^ ggg 3_ where the paradox of Gorg., 467, is 

66S.flr.,pT|ToputT)=iroAiT«^5)iopi'oiieISioXoi/, Gorg.,463D,but reaffirmed, ei iJ.iv poiiXeo-Oe (is Trai^av, el S' 0)5 airovSa^ioy; 

in Phcedr., 261 A, vfiuxaY^Y'a "i «iA Adyioi-. Cf. the defini- iJepMbJJc, 505 B, with Cfcarm., 173 £-174 B; Pep., 505 C, with 

tions of (TM^poffvn), Phcedr., 237 E. Gorg., 499 B, where Callicles is forced to admit that some 

■ 67 The Laws repeats the substance of the definition of pleasures are bad. Zellee(p. 604) thinks that Sep., 505 C, 

justice, 863E: Ti)v yi-ii rofl »vnoO .... ital eiri9i/(iii>' ev ipvxji refers to the Philetms. But the advocates of a late date 

TvpavviSa .... irapTws aSixiav npoaayopevot. Cf. 689 A B, rh fQj. tjjg philebus rightly deny any specific parallel, 

yop Kvirovittvov kox TiUp.tmv avrit (sc. T^i if^x^') ""^'P ^W" 71 Even after the Republic and Politicus, Plato in Imws, 

re icai irA^os TrdAeiit ianv. Cf. Pep., 442 A, o «)) nkilarov t^i 963 ff., approaches the problem of the "political art" and 

^X^«» 6*-^. the unity of virtue precisely in the manner of the tentative 

68Uep., 454 A; Phileb., 14 C, e«oO<ri re ital aKovcrtv; The- dialogues. There is no reason for taking seriously Socrates's 

cetet, 206 B, ixovra ^ aKovra. iraifeii"; The(etet., 167 E; dramatic bewilderment as to the "political art" in j;«tAj/- 

Sophist, 259 D ; already in Lysis, 216 A B. Cf. infra, p. 19. dem. , 292 D E, that would not apply equally to the avowal of 

69Pft(Edr., 237 C, 263, and, from a slightly different point ignorance in Laws, 963 B, or in the Politicus itself, 292 C. 

of view. Pep., 538D; Phcedo, 90C. This is largely due to a The political art, i.e., ultimate ethical and social "good," 

false conceit of knowledge, PAoedr., 237 C, which the Elen- was always a problem to Plato, as it must be to any 

chus as described in Soph., 230 B, and practised in the minor thoughtful, conscientious man {Pep., 451 A). In the Laws, 

14 The Unity of Plato's Thought 

obliterate the traces of his mental growth ? Have we not a right to expect dramatic 
illustration of so prominent a feature in the intellectual life of the time, and do we not 
find it in the Laches, Charmides, Lysis, and the corresponding parts of the Pro- 
tagoras? In brief, the Euthydemus, 277, 278; Phcedrus, 261, 262; the Thecetetus, 
167 E; the BepuUic, 454, 487 BO; the Sophist, 230 B, 251 B, 2590, and Philebus, 
20 A, 15 E, show a clear consciousness of dialectic, not merely as a method of truth, 
but as a game practiced for amusement or eristic, to purge the conceit of ignorance 
or awaken intellectual curiosity. When we find this game dramatically illustrated 
why should we assume naive unconsciousness on Plato's part? 

c) The Republic, in which Plato explicitly states his solution of these problems, 
is a marvelous achievement of mature constructive thought. But the ideas and dis- 
tinctions required for the solution itself are obvious enough, and it is absurd to affirm 
that they were beyond the reach of a thinker who was capable of composing the Pro- 
tagoras,^' the siibtle Lysis and Charmides, or the eloquent and ingenious Gorgias. 
That the highest rule of conduct must be based upon complete insight and is the 
possession of a few; that the action of the multitude is determined by habit' and 
belief" shaped under the manifold pressure of tradition and public opinion; that the 
virtues may be differently defined according as we refer them to knowledge or to 
opinion and habit; that opinion in the Athens of the Sophists and of the Peloponne- 
sian war was not guided by true philosophy, and therefore was not the " right opinion" 
which should become the fixed habit of the populace in a reformed society ; that the 
Sophists who professed to teach virtue taught at the best conformity to the desires 
and opinions of the many-headed beast, and that therefore in the proper sense virtue* 
was not taught at all at Athens;" that virtue is one regarded as knowledge, or as the 
spiritual harmony resulting from perfect self-control (443 E), but many as expressing 
the opposition of contrasted temperaments and different degrees of education; and 
that endless logomachies result from the inability of the average disputant to grasp 
these and similar distinctions " — these are reflections that might present themselves to 
any intelligent young man who had listened to Socrates, and surveyed the intellectual 
life of the time, though only the genius of Plato could construct a Eepublic from 
them. They could occur to Plato at the age of thirty or thirty-five as well as at forty 
or forty-five ; and it is extremely naive to assume that so obvious a distinction as that 
between science and opinion, familiar to every reader of Parmenides, and employed to 
bring the Meno to a plausible dramatic conclusion, was a great scientific discovery, 
marking an epoch in Plato's thought." 

964 S., as in the Republic, he finally limits himself to indi- '< iSep., 192, 193. 

eating the kind of training that will prepare the mind to f^Laws, 964 A, fitafooC &e to« epSiv koX Sirp Ttrrapa ovra 

apprehend it best. But as against the ideals of Athenian cf eo-Ti. koX iiii fie a^iov, aov Sti^avro^ us ev, TraAii' bir^ 

sophists and politicians, his beliefs were defined "already " Ttrrapa. 

in the Euthyphro, 2 C, and the Gorgias, 463 D £E., 521 D. 76 Not to dwell on the re'iemblance of Meno, 99 C, and 

'2 " One of the finest specimens of analysis in all his Apology, 220 (c/. also the Jon), why, if Plato has no dra- 

writings."— John Stcaet Mill, Dissertations and Discus- matic reserves, is opfl^ fiofa ignored in the Euthydemus? Or 

sions. Vol. IV, p. 250. is the Euthydemus, with its mature logic and its assump- 

iiPhcBdo, 82A; Sep., 522 A, 619 C ; Xauu, 966C. tion that virtue can be taught, earlier than the Meno.? 

Paul Shobey 15 

d) Lastly the structure and logic of the minor dialogues are indicative of dramatic 
design rather than of tentative inquiry. The systematic evolution of the argument 
and of the antitheses which it involves;" the emphasis laid on the very difficulties 
elucidated by the latter theory ;" the reserves and qualifications of the argument and 
the hints of dramatic purpose" — all point to Plato's possession of the clue. The argu- 
ment based on the absence from the "Socratic" dialogues of certain features of the 
longer works begs the point at issue. 

Assuming that Plato undertook to illustrate in brief dramatic discussions the 
ethical logomachies of the day, he would by hypothesis as a rule abstain from Pytha- 
gorean myths, criticism of pre-Socratic thinkers, demonstrations of immortality, psycho- 
logical or physiological digressions, and dogmatic developments of his own philosophy. 
It may be argued that such dramatic dialogues form as a whole an earlier group. It 
cannot be maintained that they mark the stages of Plato's own progress.'" The defini- 
tions of the virtues proposed in the fourth book of the Republic, interpreted by their 
context, meet the dramatic difficulties of the Laches, Charmides, Protagoras, and 
Meno. Courage is not animal fearlessness, neither is it precisely knowledge of things 
terrible and the reverse. But the courage to be expected of the masses in a reformed 
state is the conservation by disciplined feeling of the opinion about things terrible or 
not terrible inculcated by the possessors of such knowledge." I^axfypoa-vvrj is not pre- 
cisely quietness, nor doing one's own business, nor self-knowledge, though each of 
these definitions emphasizes one of the shades of meaning which Greek usage assigned 
to this "mixed mode." It is in man and state the willing acceptance by all the psychic 
faculties and the corresponding classes in the population of a harmonious scale of 

77 In the Charjnides <ru^po(rvvTi is first defined by the the problem of good and evil, aud the ultimate nature of 

quiet temperament, 159 B, then by the associated modesty, desire and the good. 

a;8<i5, 159 E, which is elsewhere its virtual synonym, Pro- ,5 Note the repeated demand that it be shown how <Tu,^po- 

tag., 322CDE; then by ri eavrov ^parr^tv, 161B, another ^^^ j^ ^ ^^^^ Charm., 159C, 161 A, 165 D, 172D, 174B, with 

rhetorical equivalent, Tim., 12 A, which, however, requires ^^^^ ^ ^y .^^^^^ p „ ^^^ ^„„^_ ,ig_ ^j^^^^ ^^^^ ^j^^^ 

an interpretation that Critias is unable to give, even though ^^^ Bepublic, it is recognized that a^<fpo<rvy,, as the mere 

assisted by a hint from Socrates (161 E). He cannot gen- p^ggi^g conditio sine qua rum of the usefulness of the active 

eralize minding one's own business, «nd distinguish (1) the ^jrtnes iK6yov 0-17^5 Sl(i.ov av Ar,. Again, cf. the association 

economic, (2) the social and political, (3) the psychic ^j ^^ ^^^^^5 npirrt^v in 161 with the division of labor, and 

division of labor; Rep., 443C. The formula is allowed to ^^^ _ 37^^^ ^^_ ^C, «3D. So in the Laches, Nicias is 

drop, and the equaUy ambiguous expression "self-kuowl- ^^j^^^^ ^^ ^^^^ ^.j^^^ ^^^ knowledge of things really terrible 

edge" is substituted (16i), which is found to involve puz- ^^^ ^j^^ reverse is not the property of any craftsman even 

zles that Critias can neither untie nor out (cf. 167 A with .^ ^^ ^^ ^^j^^ j^^^ j^ ^^^^ j^j^^g^. kno^ig^ge of final ends 

MerM, 80E; Theatet., 188 A). which he cannot define— i. c, obviously the "political art" 

In the Laches, Laches insists exclusively on the tem- ^^ ^.j^^ y^^ ^^ ^^^ 
peramental aspect of bravery which opposes it to other 

virtues, Nicias on the cognitive element which identifies it " Charm., 160 B, « yt toi)to« toO k6yov ■ the obvious de- 

with them. Laches's theory tends to show how the virtues sign of humbling Critias, 162 C D ; Charmides's disbelief in 

aremany, that of Nicias how they are one (iaws, 963 Eff.). Socrates's ignorance, 176 B. Cf. Ph<Bdr., 262 D, <it iv A 

But neither can expound his own view completely, still less eiSus to iXijWs ^poirnaiiuv iv Uyc, irapiyoi roi, iKomfTa,, 

reconcile it with the truth of- his adversary. They exemplify Laches's unfamiliarity with dialectic and the awakening 

the logomachy described in Polit., 306, 307. This is the chief effect of the Elenchus upon him ; 194 A B. 
object of the dialogue, and not the reduction of all virtue g^j^ XJebekweq says (Untersuchungen, p. 280) : "FOr 

to knowledge (ZeUer), nor the unity of virtue (Horn), nor ^^^ Verstandniss des Platonismus ist kaum ein anderer 

even the establishment of the definition i).povi,.o! <apT£pt« i„timm gefahrlicher, als der, eine Zurflckhaltung, die 

which Bonitz says is the only suggestion not disproved. pj^^.^ ^^^ methodischen Grtinden ttbte, mit einem Noch- 

In the Lysis we begin with purely verbal quibbles, pass ^ic^tsein j.u verwechselu.' 
to the suggestive antithesis of the attraction of like and 
unlike in nature and man (214, 215), and conclude with 8I ijep,, 429 C D, 442 C. 

16 The Unity op Plato's Thought 

subordination from higher to lower/" It is thus the precondition and obverse aspect 
of justice which is the fulfilment of its own function by each faculty and class — a 
higher than the economic division of labor in the soul and in society.*' These defini- 
tions are stated in terms of being rather than of doing, and Plato preferred this form 
of statement to the end/* But he is careful to add that the one includes the other 
and that the justice within the soul will express itself in just action.*^ 

3. These definitions, then, meet the chief difiiculties of the minor dialogues and fill 
their place in the literary economy of the Republic. But Plato warns us that they 
are not the final definitions of a complete philosophy."" It is not enough to define the 
virtues psychologically on the assumption that their sum is good." A final definition 
must relate virtue to, and deduce its utility from, an ultimate standard or ideal of 
good."' Such a definition is rather a regulative conception than a practical possibility. 
The Platonic Socrates is always prepared to silence by dialectic or overwhelm by his 
eloquence those who deny that "virtue" is a real good."' But a formal, positive enu- 
meration of the reasons why courage and justice are good and desirable can never be 
complete, and will always prove unedifying: "Does law so analyzed coerce you 
much ?" Plato wisely attempts nothing of the kind. He merely describes the dis- 
cipline and education"" that will enable his philosophic rulers to prove, if required, the 
coincidence of virtue and happiness, and systematically inculcate efficacious right 
opinion, thus teaching virtue and molding character and institutions in the light of a 
reasoned and unified conception of the true scope and good of individual- and public 

82 432 A, 442 D. This definition is adapted to the literary " know thyself " and '* minding one's own business." There 
machinery of the Republic. It does not estop Plato from is not one temperance or bravery, but three or four. There 
employing the word in its normal Greek sense (Bep., is no incompatibility between this view and Plato's insist- 
389 D E, wff 7rA^0€t, etc.), or from recognizing that it is a con- ence on the necessity of the definition and the final unity 
dition of virtue rather than an active virtue ; sitpra, p. 12. of virtue. If the word has many meanings, the first step 

83 Allowance once made for the literary schematism of *° rational argument is to define the one intended. And 
the four virtues, the three faculties, and the analogy be- the unity of virtue is to be sought, not in a verbal defini- 
tween the man and the state, and account once taken of *ioO' but in the unity of the moral life, the idea of good, the 
Laws, 696 C, 710, and Politicus, 306 £E., it becomes a little political art, the aKo-not (cf. infra, n. 102). The definition 
naive to complain that the distinction intended between ^^ ^ hypothesis at the beginning, or a stage in the progress 
<ra>*po<rili'7| and SucaioawTi is not clear, and a little pedantic °* ^^° argument (Charm., 163 A; Euthyphro, 9D, 11 C; 
to institute a learned philological inquiry to ascertain it. PAcedr., 237 D, ofioAovia Sejiei-oi opov, 263DE). It cannot be 

»iLawB,mk, -rriv Si ToS ipcVrov Jofav ea^ air, Kpa- ^".^""^'^^^J"^ this reason dialogues that seek a definition 

ToS<ra iv ^vxaU S.a,o»-^.„- .i;.T« i^Spa, -civ a^iM^rai tc SUat.oy i^'^" ^'"^ dialectical relativity of the definition, of course, 

7 jL ' > ' a- does not preclude Plato from arguing that his ideal of the 

moraland social life is better than that of average Athenian 

442 E, 443 A. opinion, and that the definitions which embody it are right 

86 Grote, followed by many others, denies this. But as against formulas that express some aspect of the tradi- 

that is because he persists in attributing to Plato the tional belief. 

doctrine that ethical abstractions ("mixed modes") have 87 ijep., 427E, oVat i,^:^ tV toAcv reA.o,, ivaOV 

one meaning only which can be expressed in an absolute ,j„„.. s^^ov ir, St. <ro*i r' e-rrt <al ivSpeia ,al <rii*p<ov «al 

definition; cf. supra. But, on the contrary, the very cause SiKaU. 

of the confusion, according to Plato, is that men fail to take «» r^ ■ j rn. t> /-, tn 

notice of the different meanings and sub-species covered , '°P"^-' 504 B C D, 505 A, r, toC iyaSoO iSea ,; SUaia 

by one generic term (Pfiwdr., 161, 162; Euthydem., 277, 278; ""' '' '^'"' '/""XPIo-i/ieva x/'W^H-t «" <i,(,^X.^a yiyv.rai. 

ioMe,837A; Phtie6., 12Eff.; Euthyphro, 1 B. yiith Phoedr., ^^Gwrgias; Rep.,1. 

263B, and PoUt., 285E; Polit., 306 A). Laches, Nicias, '"The "longer way," iJep., 504C, is for the guardians, 

Charmides, Critias, discuss the virtues without distin- not for us who are reading the Bepublic. See Laws, 964, 

guishing temperament, convention, habit, systematic dis- 966 C. Neglect of this point has caused much misinterpre- 

cipline, opinion, and complete insight. They are unable to tation. See Idea of Good, in " University of Chicago Clas- 

attach any precise meaning to the conventional phrases sical Studies," Vol. I, p. 190. 

Paul Shobet 17 

life. The attainment of this mastery he poetically describes as the vision of the Idea 
of Good. But it must never be forgotten that all this mysticism culminates in the 
precise and purely logical statement of 534 B C, which affirms little more than Phce- 
drus, 278 C, or than Mill when he says : "There is no knowledge, and no assurance 
of right belief, but with him who can both confute the opposite opinion and success- 
fully defend his own against confutation."' Many secondary suggestions attach them- 
selves to the phrase by association with the goodness of God, the universal cause, in 
the Timceus^ the vision of th« absolute ideas in the Phcedrus and Symposium, the 
fantastic enumeration in the Philebus (66) of the elements of "good" conceived at 
once as an ethical and a cosmical principle.'' Its chief logical and ethical significance 
for the Eepuhlic has been hopelessly misunderstood, owing to the failure to connect it 
rightly with the problem of the "good" as presented in the minor dialogues."* In 
these dialogues Socrates repeatedly tests definitions of the virtues by demanding that 
they be related to happiness, the political or royal art, or the good. A virtue by 
hypothesis must be a koXov and ayaOov.^^ The definitions proposed repeatedly break 
down because Socrates is able to instance cases in which the rule prescribed does not 
conduce to happiness — is not good."° Similarly the rhetorician, the sophist, and 
other pretenders to some supreme knowledge are confounded by Socrates's demand 
that they shall sharply discriminate their art and science from all merely instrumental 
and technical specialties which effect good or evil according as they are rightly or 
wrongly used, and show its identity with the art of arts, the art of final ends, the 
political art, the good." 

In some of the minor dialogues the negative dialectic seems to go too far, and 
Socrates makes demands that neither Platonism nor any other doctrine can meet. Thus 
in the C/iarmides the familiar expression " knowing one's self ," " knowing one's limits," 
"knowing what one can or cannot do," is made a puzzle by confounding it with the 
psychological question of self-knowledge or self-consciousness, and the fallacy or 
problem about knowing and not knowing the same thing;"* and, waiving this point, 
Socrates demands proof that knowing the things one cannot do and intrusting them 
to experts is a good — a fundamental axiom of Platonism."" The explanation is that 
the phrase, like ra iavrov irpaTTeiv above, is taken externally of adminicular and 

91 Dissertation* ond DiscussioM, Vol. IV, p. 283. the "opinion of the best" is treated as a potent cause, 

... .^, , ... Finally he identifies the idea of good with God by a sophis- 

9229 E, iya9b5 V. On the identification of the good with ^^^j interpretation of ,rap.^^>ia.a iavrv mm., 29 E) and a 

God see Idea of Good, pp. 188, 189. j^^^ constraction of (92 B) eUi.^ toO rorjTov (sc. (>fov not SeoO, 

93 Fantastic because due (1) to the wish to depress c/. 38CD). 

ilSovn to the fifth place; (2) to the neo-Platouic device of 95 jfeno, 87 D ; iocAes, 192 C, 193 D ; Protoff.,349E; Bipp. 

extending the intelligible hierarchy by the interpolation of j^i^j., 284 D ; Bep., 332, 333. 
new members between the highest and the lowest. It 96 See Idea o/ Good, pp. 200-204. 

belongstorhetoricorreligiousemotion,then,nottoPlatos ,,^„j^^^^ 282E, 290, 291C; Charm., 170B; Protag.. 

scieutilio ethics. ^^g^. Gorff., 501AB, 503D; PoUt, 289C, 293D,309C; Bep., 

94 .E. g., one hundred and fifty pages separate Zellee's ^^ jj 

treatment of the idea of good (P- 101) /'o^ ^JltTrrer »»Cy. Meno, 80E, Euthydem., 286D., The<^tet., 191 B 

of the ethical good (p. 867). In elucidation of the former J , , n 

he quotes little or nothing from the ethical dialogues and i*^- , t .,-t^x, r ™» 

cites neither PA<Edo, 99A,nor any other passage in which 99C/. Xen., Men., 4, 2, 24; JIo., 1, 11. DE; iam,732A. 

18 The Unity op Plato's Thought 

mechanical arts and sciences, not as in the Republic, with reference to the division of 
labor or function in the soul and the supreme arts of life and government. To ask 
why Critias is allowed to be baffled for lack of this distinction is to ask why Plato 
wrote short dramatic dialogues at all — why he did not incorporate the fourth book 
of the Republic in the Charmides. So in Euthydemus, 292 E, the suggestion that 
the good achieved by the possessors of the political art will be the training up of 
successors to know it is treated as a vicious circle or an infinite regress, although, when 
accompanied by the fuller explanations of the Republic, it is evidently in part the true 
Platonic doctrine.'™ And similarly in the Lysis the theory, virtually repeated in the 
Symposium, that that which is intermediate between good and evil desires the good 
as a remedy against evil, is rejected because it makes the good a mere means to an 
end.'™ But the general meaning that emerges from the airopiai of the minor dialogues, 
and the answer to them given in the Republic, is as simple as it is sound. A philo- 
sophic ethics must systematically relate its definitions and prescriptions to some con- 
sistent conception of final ends and good — be it the realization of spiritual health and 
order in a reformed society, the development of personality, the greatest happiness of 
the greatest number, the fulfilment of the will of God, the renunciation of the will to 
live, or the survival of the fittest. The statesman rises above the politician, the 
thinker and artist above the rhetorician, the true teacher above the charlatan, by his 
possession of an aim and a standard, his apprehension of a type of perfection toward 
which all his thoughts, and words, and acts converge.'"^ 

Plato's own ethical and social conceptions were thus co-ordinated and unified. 
Those of the brilliant sophists and rhetoricians who figure in his pages were not. 
They may have been very estimable and ingenious men. They could not in Plato's 
judgment be true philosophers, statesmen, or teachers of statesmen, because they 
lacked both the "idea of good" and the synoptic and unifying dialectic required for 
its systematic application in ethics and politics, and in the education of the masses to 
"virtue." This recognition of the logical significance of the idea oir good for the 
Republic and the Socratic dialogues does not commit us to an acceptance of all Plato's 
social ideals. It does not even require us to admit that the doctrine of the Republic 
really solves all the difficulties suggested by Plato's "negative dialectic." But it 
creates the strongest presumption that it was present to his mind when he wrote the 
Laches, Charmides, and Euthydemus. 

Parallel to the quest for the definition of the cardinal virtues leading to the idea 
of good is the study of friendship, love, passion, culminating in the apprehension of 
the idea of beauty at the point where it is hardly to be distinguished from the good.'"' 
No complete philosophy can ignore these things. Plato's reflections upon them have 

100 C/. Meno, 100 A, olos «<ii «AAoi/ 7roii<rai woAtTtKoi', etc. ^"iGorg., 503E, 501 C, 517, 518; Bep., 484 C, 500DE, 520Cj 

Cf.Bep., 412AB, 497CD; Laws, 950Bff.; Polit, 309D, tw ioKis, 625 E, 630 C, 688 B, 693 B, 706 A, 717 A, 733 CD, 962 A. 
S-„ ^oX.T«bv .... ^poa^^t . . . . T„- T^, paa.A«55 ^ov^Ti ,,3 ^19, 220 ; Symp., 205 D, 210, 211 ; PhcBdr., 250D ff. ; 

TOUTO avTit euTTOiet*' TOt? opCus fieTOAapovai iraidetas, wniCn, r>i. •» i. /.j -ri - t* -_» « - * ' 

however, refers partly to the lower education as well. ■•-^-j.' 

101 Cf. Lysis, 218 A, with Symp., 203 E. 

Paul Shoeey 19 

become the commonplaces of the philosophy and poetry of modern Europe: the 
strange antinomy between the love of like for like and the attraction of dissimilars in 
man and nature; the exaltation of character and mood in passionate love and friend- 
ship; the transfiguration of the passion in the love of aesthetic, moral, and intellectual 
beauty;"^ the overloading of the instinct to achieve the ends of nature — the immor- 
tality of the species."* The student of the Lysis, Phcedrus, Symposium, Republic, 
and Laws will find it impossible to fix a date at which these ideas first presented 
themselves to Plato's mind."" The mood, the treatment, the emphasis varies. Some 
of the thoughts are omitted in each dialogue, none are treated in all, and contradic- 
tions and developments may be "proved" by uncritically pressing the language and 
the imagery. But the differences between the Symposium and Phcedrus, both pre- 
sumably works of the middle period, are as noticeable as those found in any other 
works that touch on the theme. The Symposium mentions one idea, the Phcedrus 
several; the former ignores immortality and avdfivTj<rK, the latter is one of the chief 
sources for both.'" The Phcedrus ignores the thought that love is the yearning of 
the mortal for immortality, the Symposium virtually omits the doctrine of fiavia and 
enthusiasm. In the Symposium love is not a god, but a demon; in the Phcedrus he 
is ^€09 or (to escape explicit contradiction) n delov. These and other differences pre- 
sent no difficulties to a rational literary interpretation. On no reasonable theory of 
Plato's development can they signify real changes in Plato's beliefs in the interval 
between the composition of the two dialogues. 

The Lysis, though a slight Socratic dialogue, displays extreme subtlety of dialec- 
tic,"' and implies some of the most characteristic thoughts of the Symposium}'^ The 
failure to establish a formal definition, and the Socratic avowal of ignorance at the 
end prove nothing. There is a plain hint that Menexenus is an "eristic," and 
Socrates's treatment of him, so different in tone from the edifying little conversation 
with Lysis, is a mere dramatic illustration of the irXdvr] or airopCa that results from 
failure to discriminate the different meanings of an ambiguous term. Love, as the 
Phcedrus tells us, is such a term — including subordinate and contradictory species."" 
For, as the Laws say, 8B7 A, Bvo yap ovra avTo, kuI e^ dfufyolv rpirov dWo e'So? ev ovofjui 

WiZeUeT's theory thatBios is derphUosophiacheTrieb is trated, the identity of opposites as such, recurs in sub- 
a somewhat rigid and matter-of-fact interpretation of this stance in Parmen., 118 A B, and belongs to the same class as 
poetry. the quibble on «Tep<i», Euthydem., 301B; Thecstet., 190C; 

i05S3,mj,., 2OTD; Laws, 721, 773E. '^"™^?; V",lf r P^' "''° »■"""-'--'»'. P^^^f>- 13^; Par- 

losCf. Rep., 402, 403, with Symp., 2100; Bep., 490ABi ,„,^ ^^ ^^.^ ^.^^.^ *ao„o«er, etc , Symp., 203 E, which 

Laws, 688B, *pov,.r« ko. vov, ko. 6oJa ^er _.p»T05 T. LOTosLAwaKI (p. 239) thinks an important new point, in 

..( i^^Ov^ias; Sep., 499C, with I,a«,s, -IID, ora. ep<«5 O.cot ^^^^^^^ ^^^^ „f j^^ Cra*j,I««, is " already" in iym, 218 A. 

Ti- ^u,^p6vmv re «al ^«<u<av criTTiJev^a™- .vYt-^rai. Laws, ZgUer, who is " Unable to suppose" that Plato had "al- 

841 D, 636 C, Trapa ^vVtr, with Phczdr., 251 A j Laws, 837; ^^^^^ „ attained the guiding thoughts of his later system 

Gorg., 474 DE, generalization of k^6v asmSymp. ^^ g^^j^ 3,^^^^ tl^^j j„ j^^ j^^^^^ ^^^ psychological analysis 

■<" LuTOSLA WSKI Cp. 242) fails to tell us where i.ydij.vii<7tt is carried as far as is possible on a Socratic basis, but that 

is " alluded to in the speech of Aristophanes." the metaphysical explanation was revealed later. If Plato 

108 The conception of eristic, 216 A B, arguing to the ""^t tell all he knows in every dialogue, why is ivdu.v,,aK 

word, not the meaning, is as clear as it is in Rep., 454 A, "o* associated with .>t in the Symposium and RepublicT 
or Euthydem., 295 B C, and the fallacy by which it is illus- "o 263 C, 265 E. 

20 The Unity of Plato's Thought 

•KepiKa^ov iraaav airopiav koI cradTov airepyd^eTai. How familiar the two etBrj were to 
Plato appears from the almost technical use of the phrase 5t' ofioioTrjTa <f)i\iav in 
Fhcedr., 240 0. Menexenus's bewilderment is precisely on a par with that of Kleinias 
over the two meanings of iiavOdvoo in the Euthydemus.^^^ Plato is no more confused 
in the one case than in the other. The mood of the Symposium and Phcedrus is 
compatible with youth or maturity, hardly with old age. The thoughts are naturally 
not repeated in their entirety, but many of them appear in the Bepublic, or are sug- 
gested elsewhere. They are nowhere contradicted,"^ and there is no reason to doubt 
that they were essential permanent elements of Plato's criticism of life. But he was 
not always in the mood to dwell upon them. 

4. In another aspect the Platonic ethics is a polemic against hedonism. This must 
not be confounded with the modern utilitarian controversy. The modern opponent of 
utilitarianism is chiefly concerned to prove that the moral law cannot be deduced from 
experiences of utility, but has an a priori origin and requires a supernatural sanction. 
Plato does not directly discuss the origin of morality, but he explicitly disclaims the 
necessity of the sanction derived from the hope of immortality,"' affirms with great 
emphasis that the useful is the right,"* and bases all virtue on the supremacy of the 
Xoyia-TiKov or calculating reason."^ In the Protagoras Socrates is represented as 
maintaining against Protagoras by purely Benthamite arguments the identity of 
pleasure and the good."^ 

The seeming contradiction between this and the anti-hedonism of the Gorgias and 
Philebus demands explanation. It has sometimes been argued that Plato's own 
opinions on this point were reversed between the composition of the Protagoras and 
that of the Gorgias. Another explanation is that Socrates merely develops a paradox 
for the bewilderment of the Sophist. And it is true that in some parts of the dialogue 
Socrates is obviously jesting,"' and that we are warned against accepting the result 
too seriously by the reminder that both Socrates and Protagoras have maintained 

in 277E. 116 Protaff., 353-8. 

112 Grote says that in the Theoetetus the spectacle of a in 340 S. In 341 D, Protagoras, anticipating Philebxis, 
beautiful youth is not required as the indisi>ensable initia- 12E, and in language suggesting the protest against eristic 
tory stimulus to philosophy. But the Symp., 210 C, xiv ia Sophist, 259D, points out that (generic) resemblance is 
trjiiKpbi' ai/So! Ixn. and the Bep., 402 D, emphasize the unim- compatible with difference and even contrariety (cf, also 
portance of the beauty of the body as compared with that Meno, 74 D). He does not explain himself fully, however, 
of the mind. And in the same vein Socrates says, /coAbs yap and Socrates, ignoring the point, proceeds to trip him up 
el Si ©eainiTe .... 6 yifi KnAu; Xe'-yoii' xiiAds, etc., 186 E. The by a fallacious use of the principle that one thing can have 
Platonic Socrates is still the epoiTiitdt as he was in the Lysis, only one opposite. Whatever the date of the Euthydemus, 
nor can we suppose that he would ever have found the its author was aware that a word used in two senses may 
beautiful Meno as helpful an "initiatory stimulus to phUos- have two opposites, quite as early as he was capable of 
ophy" as the snub-nosed Theeetetus. writing the Protagoras. The passage is merely a dramatic 

113 Bep., 363 B C D, 367 E, 612 B C. The Oorgias does not illustration of Socrates's superiority in the same of ques- 
difier herein from the Bepublic, as Eitohie {p. 156) seems to tio° ^^^ answer. Again in 350 B-351 A, when it is argued 
think. The argument is complete without the myth, and the "i^* bravery ia knowledge because knowledge imparts con- 
phrases at the end about living justly in order to prepare Mence, Protagoras points out that we cannot convert the 
for the judgment of Minos prove no more than the iva of universal afarmative proposition, "all bravery is confi- 
Bep. 621 C. dence," and distinguishes as bravery the confidence that 

arises from nature and training. Though not a match for 
lUKaKov, Bep., 457 B. Socrates, Protagoras is a far better reasoner than Laches 

ii'Eep., 440E, 571C, 605B. or Nicias, and again Socrates refutes him only by taking 

Paul Shobey 21 

theses incompatible with the positions from which they started."* But the full expla- 
nation lies deeper. In the Republic Plato undertakes to demonstrate the intrinsic 
desirability of virtue against two forms of disbelief — the explicit skepticism of the 
cynic, who affirms that natural justice is the advantage of the stronger and human 
justice an artificial convention, and the unfaith of the ordinary man, who virtually 
admits this theory by commending justice solely on external and prudential grounds."' 
The CalLicles of the Gorgias represents the former view, Gorgias himself and (less 
obviously) Protagoras the latter. Like other Sophists, he is the embodiment of average 
public opinion which his teaching reproduces."" He himself says that all men teach 
virtue. He modestly claims at the most only to teach it a little more effectively and 
persuasively than the layman.™ Plato would .admit both assertions, with the reserva- 
tion that the virtue so- taught hardly deserves the name, and that the teaching is 
neither systematic nor philosophical. 

The molding power of public opinion, operating through countless social and 
educative agencies, is admirably depicted in the myth attributed to Protagoras, the 
main thought of which is repeated in the Bepublic.^'' There, however, the philosophic 
rulers are to employ this irresistible force for the inculcation, not of average Greek 
opinion, but of Platonic virtue. The Protagoras dramatically illustrates the dialectic 
incapacity and philosophic" superficiality of the great popular teacher. His ethical 
teaching is spiritually and logically on a level with the precepts of the worthy sires 
and guardians satirized by Adeimantus."^ However unlike in temper and practical 
effect, it is philosophically akin to the individual hedonism of Callicles and Thrasy- 
machus who reject all morality as an unreal convention. Protagoras is naturally 
unaware of this. Like the populace, he recoils from the naked exposition of the 
principles implied in his preaching and practice. He accepts the terminology of indi- 
vidual hedonism only under compulsion of Socrates's superior dialectic. But Socrates's 
explicit challenge to him and the assembled Sophists to name any other final good 
than '^BovT) is a proof that one of Plato's objects was to identify the Sophistic ethics 
with hedonism.'^* But neither this nor the demonstration of Protagoras's inability 
to cope with Socrates in dialectic exhausts the significance of the dialogue. 

Plato, however reluctantly, always recognized a certain measure of truth in the 
Benthamite analysis here attributed to Socrates. He knew that " act we must in 
pursuance of that which (we think) will give us most pleasure " Even the Gorgias 
contains phrases of utilitarian, if not hedonistic, implication.'^^ The Eudsemonism of 

np a new line of argument— the identity of pleasure and ^^^Bep., 362 E S. Of. Zelleb, p. 603, n. 1. 

good, and the consequent unity of the virtues in the 120 ijep., 492 ff. 121 Prof as., 328 B. 

"measuringart." Platoof conrsewasawarehere,andinthe 

„ , , ,° , u ii. I ■ „■„„! „»! , 12JE1TOHIE (p. 156) says: "The argument of the Sophist 

SMtAt/pAro (12), and everywhere, that a universal amrma- • ^ „ ^ j v m i n ^ 

ji«.c./ti,>//wv I o;, o J , „ .J. „. tu. Protagoras .... is now fully accepted by Plato," etc., as 

tive cannot be directly converted. But it is a part of the ., „, f , ,, . , i iv ^ * 

, " ,. , iv i -D i. „ , ,1. 1.1 „„!,.>„„„„ if Plato was not the author of the Pr-otofforas. 

scheme of the dialogue that Protagoras should make some 

good points, though defeated in the end. And Socrates is 123 Rep., 362 E. 12< 354 D, 358 A. 

baffled in or fails to complete other proofs of the unity of 125 499 D. Ritchie (p. 155) strangely says that in the 

virtue, and so is driven to rely on the proof from hedonism, Eepublic Plato recognizes, in marked advance upon the 

which is the chief feature of the dialogue. position of the Gorgias, that there are good pleasures as 

1 18 Protag. , 361 . well as bad ! 

22 The Unity of Plato's Thought 

the Republic has often been pointed out,'^° and in the Laws Plato explicitly declares, 
in language recalling that of the Protagoras, that it is not in human nature to pursue 
any course of action that does not promise a favorable balance of pleasure.'" But the 
inference which he draws is not that it is safe or desirable to proclaim that pleasure 
is the good, but that it is necessary to demonstrate that the good — the virtuous life — 
is the most pleasurable. 

To a Benthamite this will seem a purely verbal or rhetorical distinction. And 
Aristotle himself hints that Plato's aversion to the name of pleasure cast a suspicion 
of unreality over his ethical teaching."' But Plato is not alone in his aversion to the 
word. Matthew Arnold acknowledges a similar feeling. And Jowett, in his admirable 
introduction to the Philebus, has once for all set forth the considerations by which 
many clear-headed modern thinkers, who perfectly understand the utilitarian logic and 
accept whatever is true in its psychology, are nevertheless moved to reject its language. 
The Greek word '^Sovq is much more closely associated with a low view of happiness 
than the English word "pleasure;" and Plato had, or thought that he had, much 
stronger reasons than the moderns have, for identifying hedonism with the negation of 
all moral principle. 

The Gorgias and Philebus nowhere explicitly contradict the thesis of the Pro- 
tagoras that a preponderance of pleasure, rightly estimated and abstracted from all 
evil consequences, is good.'^' The doctrine which they combat is the unqualified iden- 
tification of pleasure and good, coupled with the affirmation that true happiness is to 
be sought by developing and gratifying the appetite for the pleasures of sense and 
ambition.'** Plato represents Callicles and Philebus as unable or unwilling to limit 
these propositions even by the qualifications of the Protagoras. ^^^ It is he, not they, 
who introduces the distinction of pure and impure,"^ true and illusive,"' wholesome 
and unwholesome,'" necessary and unnecessary pleasures.'^* The modern critic may 
object that Plato was not justified in attributing to any contemporaries either this 
dialectical incapacity or this cynical effrontery. Plato thought otherwise. It is a 
question of historical evidence. But it is not legitimate to attribute to the Callicles 
and the Philebus of the dialogues the utilitarianism of Grote or John Stuart Mill, or 
even that of the Protagoras, and so convict Plato of self-contradiction.'^° 

With these remarks we may dismiss so much of the Gorgias and Philebus as is 
merely dialectical, dramatic, or rhetorical, directed against the crudest form of hedonism 
which Plato chooses to bring upon the stage before grappling with the problem in 

126 357 B, riSovai otrat a^XajSet? goods per se,* 457 B, 458 E, and the explanation that some painful goods are medicliial 
581 B (with Laws, 732 E), nij ort irpbt to iciXAiov itaX alo-xioi' ?i)i» (354 A = Sep., 351 C), and is checked by the calculus of all 
iu.i)5e TO x^^pov (cat afj-cLvov, ctAAa jrph^ auTo TO ^Stoi' KaX a^vTTOT- consequences, -all of which is ignored by Callicles and 
tpov. Philebus. 

127 Laws, 733, 734 ; cf. 663 A. 128 Mh. nic, X, 1. 132 pMleb., 51, 52. 133 IMd., 36 C ff. 
l29PAiJe6., 60AB, is verbally a direct contradiction of mibid., 41A; Gorg., 499DE. 135 iJep., 558 D. 

^^■' ' 136 Plato, as Jowett says, is "playing both sides of the 

130 Gmg., 495 A, 492 D E ; Phileb., 12 A, 12 D, 27 E. game .... but it is not necessary in order to understand 

131 The verbal identification of ^Sovt) and ayiL&hv in 355 him that we should discuss the fairness of his modes of 
has been preceded by such phrases as ita9' h ijiia iariv, 351 C, proceeding." 

Paul Shorey 


earnest."' The real arguments which he employs, not so much to refute the thesis of 
the Protagoras as to limit its practicable application and justify his repudiation of its 
terminology, may be summed up as follows: The distinction between good and 
bad pleasures once admitted, the statement that pleasure as such is the good, becomes 
an unreal abstraction."' The reality is specific kinds of pleasure and the principle of 
distinction, whether intelligence, measure, or the will to obey the "opinion of the 
best," "' becomes more important than the bare name of pleasure, and more nearly 
allied to the good."" The "measuring art" postulated in the Protagoras is impracti- 
cable. Pleasure and pain are, like confidence and fear, foolish counselors;'" either 
deprives the mind of the sanity required for a just estimate.'" No scale of human 
judgment can be trusted to weigh the present against the future, and make allowance 
for all the illusions of memory, hope, and contrast."' The most intense pleasures and 
pains are associated with a diseased condition of mind and body.'** And the habit of 
pursuing pleasure, of thinking and speaking of it as the good, tends to make the world 
of sense seem more real than that of thought and spirit.'*^ The contrary is the truth. 
The world of sense is a pale reflex of the world of ideas,'*' and the pleasures of 
sense are inherently unreal, illusory, and deceptive, and may in sound logic be termed 
false, as fairly as the erroneous opinions that accompany them.'*' They are false 
becaiise composed of hopes and imaginations not destined to be fulfilled;'** false, 
because exaggerated by the illusions of distance in time or contrast;'*' false, because 

i37pftiJe6., 55AB, and Gforp., 495 C, 499 B, show that the 
arguments of Gorrg.., 495 C-499 B, are, in the main, a con- 
scious dialectical sport. I recur to this point so often be- 
cause the Gorgias and the first book of the Republic are the 
chief source of the opinion, widely spread by Qrote, Mill, 
and Sidgwick, that Plato is a magnificent preacher, but 
often a weak reasoner. Cf. Mill, Diss, and Discuss,, IV, 
291 : " This great dialogue, full of just thoughts and fine 
observations on human nature, is, in mere argument, one 
of the weakest of Plato's works." Cf. Idea of Good, pp. 

iisPhileb., 12 DE. In answer to the question, Trii yap 
rjSoviq ye ijSovg fj-Tj ovx, ofioioTarov av elij ; Socrates shows that 
generic (verbal) identity is compatible with specific diflfer- 
ence or even opposition, a logical principle "already" 
glanced at in the Protag., 331 D, with the same illustration 
of (icAav and Ktvicov. LuTOSLAWSEI, p. 467, misunderstands 
13 A, TovT^ Tw \6yta firi miTTeue, T<5 iravra Ta efaxTiiiraTa ev 
irotovvTi — "we need not attempt a reconciliation of all con- 
tradictions I " 

139 PAcedr., 237 D, ifk^vros .... iiriOviiia. ^fioi'ui' .... 
eiriKTijTos ho^ay e0ie/xef 7) toO apiffTOv. Cf, Laws, 644 D, 645 A. 
Phcedo, 99 A, un-b 5o^t)5 t^epoixeva ToG ^eATiVTOV. 

^^Phileb., 64C, Ti . , . . iiaXiar* atnor elvcu fiofeiej/ av ToO irafft yeyovevai irpocr^iA^ T»)»' TotavTijc Biaffeaw • with 
the context. 

1*1 Cf. Tim., 69 D, with Laws, 644 C. 

UiB^., 402E; Phileb., 63D; Phcedo, 66B. 

"3Cy. Phileb., 41Eff., with Protag., 356, 357; Gorgias, 
500 A, 3p' oiv ffavTos avSpoi eariv efcAef avdai, etc. Laws, 663 B, 

iTKorofiii'iai' Se to iroppiaOev opiafievov na<ri re wf tnoi elirelv . . . . 
irapexei, and the rhetorical repudiation of the whole hedo- 
nistic calculus, Phcedo, 69 A B. 

^^ Phileb,, 45 B-E, ev tcvi irovripiq ^x^^ '*''' '^**" aiotxaTK 
.... fliyttrral nev riSovai, etc. 

1*5 Cf. Phcedo, 83 D, with James's Psychology, Vol. 11, p. 
306 : " Among all sensations, the most belief-compelling are 
those productive of pleasure or pain." 

i«ieep., 509, 510, 514 a., the allegory of the cave. 

H' Phileb., 36 C ff. As Berkeley and Huxley argue from 
the subjectivity of pain to that of sensations and ideas ; as 
Epicurus proceeds from the reality of pain to that of the 
other secondary qualities ; so, reversing the order, Plato 
infers the falsity of pleasures and pains from that of the 
associated perceptions and beliefs. Qrote, Jowett, Horn, 
and others pronounce the whole train of reasoning feilla- 
cious. But it is to be observed : (1) that their objections 
as usual are anticipated by Plato (Phileb., 38 A), who has a 
right to use his own terminology provided his meaning is 
unambiguous (Charmides, 163 D) ; (2) that the epithet 
" false " is used either with reference to a postulated objec- 
tive judgment of life as a whole, or as a mere rhetorical 
expression of the disdain or pity felt by an onlooker. In 
the first sense it is justified by the argument, in the second 
by the usage of the poets— falsa licet cupidus deponat 
gaudia livor (Propert,, 1, 8, 29) ; (3) having demonstrated 
against Sophistic negations that ^evSi^i applies to £6^a, 
Plato was naturally tempted to extend it to iSoi-^. 

i*sphileb,, 39E, 40C. Cf. "we are aU imaginative, for 
images are the brood of desire " (George Eliot). 

wibid., 41, 42B; Laws, 663 B. 

24 The Unity of Plato's Thought 

■what we mistake for positive pleasure is usually the neutral state, the absence of 
uneasiness, the cessation of pain.™ 

This doctrine of the negativity of what men call pleasure is the fundamental basis 
of Plato's ethics, as it is of Schopenhauer's. On this, in the last instance, rests his 
refutation of hedonism, and, as we shall see, his demonstration that virtue and happi- 
ness are one.'" Sensuous pleasures are in their nature impure and illusory. They are 
preconditioned by, and mixed with, desire, want, pain. "Surgit amari aliquid" is ever 
true of them. They are the relief of an uneasiness, the scratching of an itch, the 
filling of a vacuum.'^^ To treat them as real, or to make them one's aim (except so 
far as our human estate requires), is to seek happiness in a process rather than a 
state,'" in becoming rather than in being. It is to bind one's self to the wheel of 
Ixion and pour water into the bottomless jar of the Danaids.'^* Far happier, far more 
pleasurable, is the life that consistently aims at few and calm pleasures, to which the 
sensualist would hardly give the name, a life which he would regard as torpor or 

Both the physiology and the psychology of this doctrine have been impugned. 
It has been argued that, up to the point of fatigue, the action of healthy nerves involves 
no pain, and must yield a surplus of positive sensuous pleasure. It is urged that the 
present uneasiness of appetite is normally more than counterbalanced by the anticipa- 
tion of immediate satisfaction. Such arguments will carry no weight with those who 
accept Plato's main contention, that the satisfactions of sense and ambition, however 
inevitable, have no real worth, and that to seek our true life in them is to weave and 
unweave the futile web of Penelope. Whatever qualifications modern psychology may 
attach to the doctrine, it is the logical basis of Plato's ethics. The unfeigned recogni- 
tion of the inherent worthlessness of the lower pleasures removes at once the motive 
and lures to evil."" It is the chief link in the proof that virtue is happiness. It 
insures the domination of reason over feeling and appetite. It molds man into that 
likeness to the divine pattern which is Plato's favorite expression for the ethical ideal,'" 
for the divine life knows neither pleasure nor pain."* It is the serious argument that 

iMPfetleft., 42Cff. ; Bep., 583D. some moderns, that pleasure is not strictly = /ciVtjo-ij is 

151 The argument that pleasure is yivvn.^, not oi(ria, is beside the point, 
not, as Zellee says (p. 604), the nerve of the proof. It is ^^*Gorg,, 493B, TeTpijiueVo! iriSos, etc.; PhcBdo, 84A, 

obviously, as the language of 53 C implies, one of those half- av^n/roi' epyov .... IIijceAoinjs — ttrTov, Gorfir,, 507E; Phileb.^ 

serious metaphysical and rhetorical confirmations used to 54 E. 

make a strong case where Plato's feelings are enlisted. It 155 pftcedo, 64 B ; Oorg., 492E; Phileb., 54E, itai' cjao-i f?./ 

does not occur explicitly in the Republic which speaks, how- oi^ i„ aef aaSai, etc. In Laws, 733, 734 B, the hedonistic calcu- 

ever, of pleasure as KirTjo-i!, 583 E. lus of the Protagoras is retained, but is applied not directly 

i52"Already" in the Gorgias, 493E, 494C, and the to the individual acts, but to types of life. The life of 

Pftoedrits, 258 E,«»'irpoXvin)flV<" 8"% *"?5e VSiji'ai, etc. ; JJep., moderate pleasures is a priori the more pleasurable 

584 A B. It has even been argued that the Pftosdnts passage because it necessarily yields a more favorable balance 

takes for granted the fuller discussion of the Philebus (W. than the life of intense pleasures. 
H. Thompson, PAcEdrits, ad !oc.). And why not? Anything issp^cedo, 66Ci Pep., 586AB 588. 

may be argued if the dialogues are supposed to grow out of . i •! , . 

one another and not out of Plato's mind. '" Thecetet., 176 B £E. ; Laws, 716 D, 728 A B ; Bep., 352 B, 

l53PAiIeft., 53Cff.; 54EvirtuaUy = Gor(7., 493E. The 612^; PAiiet., 391E. 
Uteral-minded objection of Aeistotlb, Eth. Nic, X, 4, and 158 Phileb,, 33 B. 

Paul Shobey 25 

explains Plato's repudiation of the hedonistic formulas of the Protagoras, and justifies 
the noble anti-hedonistic rhetoric of the Gorgias, the Phcedo, and the Philebus.^^ 

4. Plato's insistence on the necessity of proving the coincidence of virtue and 
happiness marks another difference between him and modern writers. The question 
is rarely put in the forefront of modern ethical discussion, except for the polemical 
purpose of proving that an opponent's philosophy supplies no basis or sanction for 
morality. The majority of modern ethical writers relegate the problem to a digression 
or a footnote. They are content to establish a "general tendency" or "strong proba- 
bility." Or they frankly admit that while everybody would be glad if the proposition 
could be proved, it is not susceptible of mathematical demonstration. But this was 
not enough for Plato. His own faith was adamantine.'^ He was as certain that hap- 
piness is inseparable from virtue as of the existence'" of the Island of Crete. 
Even if it were only a probability, he would not permit it to be impugned in a well- 
ordered state. "'^ Just how much positively immoral and cynical philosophy was cur- 
rent in Plato's day is, as we have seen, a disputed historical question. But Plato 
himself was haunted by the thought of the unscrupulous skeptic who sought to justify 
his own practice by appeals to the law of nature or theories of the origin of justice in a 
conspiracy of the weak against the strong.'"' His imagination was beset by the picture 
of some brilliant young Alcibiades standing at the crossways of life and debating in 
his mind whether his best chance of happiness lay in accepting the conventional moral 
law that serves to police the vulgar or in giving rein to the instincts and appetites of 
his own stronger nature.'" To confute the one, to convince the other, became to him 
the main problem of moral philosophy. It is a chief duty of the rulers in the Mepicblic 
and the Laws, and the Socrates of the dialogues is at all times ready and equipped to 
undertake it. 

Plato is not always overnice in the arguments by which the skeptic is refuted. It 
is enough that the "wicked" should not have the best of the argument.'*^ Socrates in 
the first instance puts forth just enough dialectical strength to baffle a Callicles or a 
Thrasymachus."" This, as we have seen, is the quality of much of the argument of 
the Gorgias,^^^ though it is intermingled with hints of deeper things, and supplemented 

159 Gtwff., 507, 512, 513 ; Phmdo, 69 ; PMieft., 66 A ; Bep., 580 B. tration of the game of question and answer. Thrasymachns 

leoGorgios, 509A; Bep., 360B, 6i&A. sets np the thesis, oi aSiKoi ^pivtiiM koX iyaSoi, and Socrates 

101 i^aws 662 B. forces him to contradict himself. ZeUer (p. 752) lists it 

i62iSep.,'392AB; Z,at««,663B, mSavrfs y\ ei ^r,Uv irtpov, among Plato's fallacies. 

Trpos TO Tira i9ikv.v irjv rhf ocrioi' koX Si/taiov jSioi-. '6' Strictly Speaking, Soorates's dialectic is employed 

l63iJep., 358, 359, 365; Oorg., 483 ft. Cf. Bep., 358 C, merely to force from Callicles the admission that some,^ivo, Ta Sra; Protoff., 333 C, ineX noXl^oi yi ««a.; Pleasures are bad (M9BC; cf. Bep., 505C). From this 

Euthvdem.,219B,i^:«,yipi. t« ,,,.1. a^4>^apr,r^<r,..; Phileb., POint the argument, abandoning ethical theory discusses 

66 E i Gorg., 511 B ; Laios, 889 D E, with Theostet, 177 C D. social and political ideals at Athens. Good is treated as 

distinct from "pleasure, as it IS in PAcedr., 239 C. But the 

16* iJep., 365 B ; Gorg., 510 D ; Laws, 662 E. question whether it may not ultimately prove to be the 

^^Thecetet, 176 CD; 177 8, kAi i, priropiKri Ueivri jrm favorable balance of pleasure (Protaff.) is not raised. The 

oirofiapaiVeTm. The whole passage is a description of the crude identification of the terms is rejected for reasons 

Gorgias. Cf., 527 A, vvv U op^t, ort rpeli ovTei vjieit, oiirep gyji held valid in the Philebus. Cf. Phileb., 55 B, with 

(ro^urciToi eirre rSiv vvv 'EK\-^viav .... oiiK i^ert diroSiiJai, etc. Gorg., 498 C. There is no contradiction. The three dia- 

Laws, 907 C, ri "OTe \6yois JiyuvTOLi KpaTovvret, etc. logues, differing in mood, are logically consistent and sup- 

16«E. g., the argument in Bep., 349, 350, is a mere iUus- plement one another. 

26 The Unity of Plato's Thought 

by noble eloquence. In the Republic, however, Plato undertakes not only to confute 
and silence, but to convince.'*' The real ground of conviction is the total underlying 
conception of the true nature, harmony, health, and consequent happiness of the soul. 

But the formal proof is summed up in the ninth book in three arguments which, 
as Plato repeatedly tells us, constitute the framework of the whole design."" To these, 
in form at least, all other interests of the book are subordinate — the construction of 
the ideal state, the higher philosophical education, the idea of good, the character- 
sketches of degenerate types. The first argument is based on the comparison of the 
individual and the state which runs through the entire work from the second to the 
ninth book. It takes two forms : (1) That of a mere external analogy. As the hap- 
piness of the ideal state is to the misery of the ochlocracy or the tyranny, so is the 
happiness of the well-governed just soul to the wretchedness of the man whose soul is 
the prey of a mob of appetites, or the slave of a ruling passion.'™ (2) The force of 
this external analogy is derived wholly from the psychological truth that it embodies. 
Unity or factious division, the sovereignty of reason, or the usurpations of passion 
and appetite, harmony or discord, health or disease, as used of the soul, are more 
than mere figures of speech ; they are the exact expression of inevitable alternatives 
resting on indisputable psychological facts. The dominance of the higher reason over 
disciplined emotion and controlled appetite is the sole and effective condition at once 
of the unity, harmony, and health of spiritual life which is happiness, and of the 
unswerving fulfilment of obligation which is the external manifestation of justice and 
virtue.'" To ask whether happiness is compatible with a diseased soul is still more 
absurd than to expect it to dwell in a diseased body."^ 

The second argument is very brief, and Plato is probably aware that at the best 
it commands assent rather than inspires conviction.'" The three faculties of the soul, 
taken abstractly, yield three types of pleasure — the pleasures of pure intelligence, of 
ambition, and of appetite. Plato assumes that the pleasures of intelligence belong to 
the man in whom the intellect directed toward the good controls the other faculties. 
In other words, he takes for granted the coincidence on the highest plane of intellect 
and virtue which he found in Socrates and which the education of the Republic secures 
in the guardians.'" Now, the advocate of the intellectual and virtuous life has neces- 
sarily had some experience of the pleasures associated with gratified ambition and 
appetite. The ambitious man and the sensuous man know little or nothing of the 
higher order of pleasure. The preference of the "intellectual" for his own type of 
pleasure must be ratified as based on a completer experience. It would be a waste of 
time to cavil on minor fallacies or rhetorical exaggerations with which Plato burdens 
the argument in his eagerness to make a strong case.'" The argument itself is familiar 

168 Bep., 357 A B, 358 B, 367 A B, 367 E. "3 Rep., 580 D B. "* Of. supra, p. 11. 

169369AB, 392AB, 427D, 445A, 544A. "SGrote and Mill object that this argument, even if 

1105760 S 171 442 E conclusiTe, is addressed to the wrong point, because the 

life supposed is not that of the simple, just man, but that 

"2 445 A, 591 B, 589 E ; Gorg., 512 A, 479 B ; "already" in of the philosopher But the case of the simple just man is 

Crito, 47 D E. met by the main arguments drawn from the order, har- 

Paul Shobet 27 

enough through its acceptance in substance by John Stuart Mill ; who, however, seems 
to think Plato's use of it fallacious. It has been rejected as a fallacy on the ground 
that pleasure is not an objective measurable entity, but a relative individual feeling. 
Again at the limits of human thought we are confronted by an alternative the terms 
of which it is impossible to realize distinctly. Is it better to be a completely con- 
tented pig than a man? But if we waive the claim that the argument is an absolute 
proof, and turn from these unreal abstractions to the facts of life, what Plato affirms is 
simply that it is more pleasurable in the end to develop and foster the capacity for 
the "higher" pleasures than that for the lower, as is shown by the judgment of those 
who have experienced both. In this less absolute form the argument leans for support 
on that which precedes, and still more on that which follows it. 

In the third place, the lower pleasures as compared with the higher are illusory, 
unreal, and impermanent, and they tend to destroy the healthy balance of faculties 
which is the condition of all true pleasure.''^ This is a repetition or anticipation"' 
of the theory of the negativity of pleasure which we have already met in the polemic 
against hedonism. 

This completes our sketch of the Platonic ethics. The rest is exhortation, inspi- 
ration, myth, things ovk a-qhearepa aicoveiv, but not within the scope of the present 
study, nor indeed reproducible in any study. For the ethical and religious spirit that 
informs every page of Plato we must go to the master himself. 


Plato's theory of ideas is (1) primarily a realistic way of speaking of the univer- 
sal; (2) a poetic and mythical extension of this realistic language, by which the uni- 
versal is treated, not only as a thing, but as a thing of beauty and object of desire and 
aspiration ; (3) in relation to metaphysics, it is the definite and positive assertion that 
the substantive essences, or rather the objective correlates, of general notions consti- 
tute the ultimate ontological xinits of reality to which psychological and logical 
analysis refer us as the only escape from a Heraclitean or Protagorean philosophy of 
pure relativity. In the first sense the ideas occur throughout the dialogues. It is 
irrational to look for the other forms of the doctrine except when the argument natu- 
rally leads up to them. A Kantian does not expatiate upon the Ding-an-sich in an 

mony, and health of the soul, and from the analysis of proximate to these types. And the statement of the argn- 
pleasure. Here Plato is renewing the debate between the ment in the Laws applies to the simple just man, 663 C, 
" philosopher," the sensualist, and the politician begun in ri a8i«a . . . . U iiiv iSUov «al naicoS iovroC Oempoviitva ^Sia, 
the Gorgias. He is indulging his feelings in a demonstra- etc., . . . . -niv S' aAijSeini' t^s /tpiViio! noripnv KvpiuTepav elvai 
tion that in the Athens of his day the " philosophic " life j/Hiiefi nirepa t>|i/ t^s xe'P'X'ot <pvxv^ v Ttji- t^s ^cKtCovik. 
is a higher and happier type than the life of the politician ,,5 ^ ^^ B-586 C. 
or the sensualist; and he holds that no real reform is pos- 
sible until men can be found who approach political life as i" Zeller thinks it a rfeumfi of the fuller treatment of 
anecessary, not a desirable, thing, condescending to it from the Philebus. Those who put the Philebus late regard it 
a life which they feel to be higher and more pleasurable as a preliminary sketch. The Philebus is probably late, 
(cf. Sep., 521 B) . The form of the argument of the Republic as Mill aSSrmed before Sprachstatistik was conceived, 
is determined by the purpose of contrasting the extreme But the psychology of pleasure in the two dialogues sup- 
types of the virtuous philosopher and the finished tyrant. plies no evidence. Cf. infra, "Plato's Psychology," and 
But it applies to other men in proportion as they ap- Part II. 

28 The Unity of Plato's Thought 

essay on universal peace. Plato discussed many topics that did not require embellish- 
ment by the mythical description of the idea as type, or the explicit reaffirmation of 
the idea as noumenon. And the apparent absence of either from a given dialogue 
proves nothing. 

Plato's fearless and consistent realism is so repugnant to "common sense" that 
modem critics either take it as proof of the naivete, not to say childishness, of his 
thought, or extenuate the paradox by arguing that he could not have meant it 
seriously and must have abandoned or modified the doctrine in his maturer works. 
All such interpretations spring from a failure to grasp the real character of the meta- 
physical problem and the historical conditions that made Plato adopt and cling to this 
solution. Prom Heraclitus to John Stuart Mill human thought has always faced the 
alternative of positing an inexplicable and paradoxical noumenon, or accepting the 
"flowing philosophy." No system can escape the dilemma. Plato from his youth up 
was alternately fascinated and repelled by the philosophy of Heraclitus. No other 
writer has described so vividly as he the reign of relativity and change in the world 
of phenomena."* Only by affirming a noumenon could he escape Heracliteanism as 
the ultimate account of (1) being, and (2) cognition."' He chose or found this noume- 
non in the hypostatized concepts of the human mind, the objects of Socratic inquiry, 
the postulates of the logic he was trying to evolve from the muddle of contemporary 
dialectic, the realities of the world of thought so much more vivid to him than the 
world of sense.'*" This is the account of the matter given by Aristotle'*' and con- 
firmed by the dialogues. Except in purely mythical passages, Plato does not attempt 
to describe the ideas any more than Kant describes the Ding-an-sich or Spencer the 
"Unknowable." He does not tell us what they are, but that they are. And the diffi- 
culties, clearly recognized by Plato, which attach to the doctrine thus rightly limited, 
are precisely those that confront any philosophy that assumes an absolute. 

Plato's particular selection of the hypostatized concept for his absolute seems 
more paradoxical only because, from the common-sense point of view of a convenient 
but inconsistent conceptualism, we ignore the real philosophical alternative of consist- 
ent nominalism or consistent realism, and forget the historical conditions that forced 
Plato to make his choice. Realism was for Plato not merely the only metaphysical 
alternative to Protagorean relativity ; it was the only practicable way of affirming the 
validity of universals and abstract thought. The psychology and logic of modern 
nominalism as gradually worked out by Locke, Berkeley, John Stuart Mill, and 
Taine, did not exist. The modern flowing philosopher can give a plausible account of 

178 Symp., 207D E; Tim., 43 BC, 44 A B, 52 E, 69 C D; concepts ideas (which he didl) if his starting-point had 

ThecBtet., 156 3. been the hypostatization of the concept, and (which is 

1'9 Cratt/l; 439, 440 ; ThetBtet., 179 B., 185, 186 ; Tim., 27 D, partly true) that he would not have put forth the paradox 

28A, 49Dff., 51 BC. Less directly pertinent are Soph., at all if he had not felt the necessity of positing some 

249B; Cratj/J.,386; PAt!e6., 58E, with iJep., 533B. reality beyond the world of sense. This last Apelt confirms 

I80I do not mean that Plato said: "Go to, I need -a by Met, 10406, 27, which, however, proves nothing for Plato, 

rwumenon, I wUl hypostatize the Socratic concepts," as it merely states a favorite thought of Aristotle, 
which a malicious critic might infer from Apelt's argu- "i Met, 1, 6, 987a, 29 ff., 10866. 

ment {Beitr&ge, pp. 81-3), that Plato would have made all 

Paul Shoeey 29 

the universal, recognizes the general term as a convenient algebraic symbol, and so 
accepts the old logic as a practical working instrument of thought. But in Plato's 
time the old logic was still to be created, and the cruder forms of nominalism and 
relativity which he combated blocked the way by captious objections to the normal 
and necessary use of general terms.'*" The theory of ideas, then, often appears to be 
mainly, if not merely, an affirmation of the concept apart from expUcit insistence on 
any theory of its psychological or ontological nature.'*' But the main issue is 
unaffected by this fact. Even if he had been acquainted with the analysis of Mill 
and Taine,'** Plato would have continued to ask : Are the good and the beautiful and 
similar essences something or nothing?"' Can everything in the idea be explained as 
the natural product of remembered and associated sensations?"' Is not man's power 
of abstraction something different in kind from any faculty possessed by the brute?"' 
Not all the refinements of the new psychology can disguise the fact that the one 
alternative commits us to the " flowing philosophers," the other to some form of Pla- 
tonism. For the answer that the "good" and the "beautiful" are only concepts of 
the mind is an evasion which commends itself to common-sense, but which will satisfy 
no serious thinker. If these concepts are the subjective correlates of objective reali- 
ties, we return to the Platonic idea — for Plato, it must be remembered, does not say 
what the ideas are, but only that they are in some sense objective and real."* If the 
concepts are the natural products of casual associations, accidental eddies in the 
stream of sense, the "flowing philosophy" receives us again."' Moreover, though this 

1S2 PJiileb,, 14 D, ir^oBpa rots Adyots itiiroSia; 15 A, 13DE; iiriSiv ayaBov elvai nrjBe Kokov .... wAiji' €v ^v\^, Parmcn.^ 

Parmen., 135C; SppA., 251BC; Thecetet, 157AB, 167A, 130B, Kal koXov koX iynSov, etc. 
180D; Euthydem., 301 A and passim. leepftcecto, 96B; Themtet., 156, 157, 184D, el noxkai t..„ 

I83i?ep«6., 596A; Phcedr,^ 249 B, though immediately iv rnjiZv toairep iv Bovpeioi^ 'innots aiaOjjaeti iyKddTjvTtu, aWa t^r) 

followed by avajUKija-lf ; Philebus^ 16 D, and all passages that dy fiiav rtca iSiav, elre \livxriv eire o rt fiet KoXelv navTa ravTa 

describe the true method of generalization and division — JwreiVei. Tim., 51 C, ri ravra oiirep ical /SMirojxo' (c/. oiirtp 

Phoedr., 265, 266, 270 D ; SopA., 226 C, 235 C, 253; Poitt., 285A; ipfev, Bep., 515B; Parmen., 130 D), So-a re aAAn «ia toO 

CTCt/tyl., 424 C; Laws, 894 A, 965 C. <rw/AaT05 altrdavoiifda, p.6va. earl ToiavTijv exovTa dXij^eiai', 

18* To Mill (Diss, and Discuss., IV, p. 300) the Platonic iSTPhcedr., 249 B, Set yip iv9pmirov (vviivai Kar' elSos K^yo- 

ideas " are only interesting as the first efforts of original ii.tvov, U iro>Mtv iov aiaBrjacciv tii Iv Koyi.aii.ii ^vraipoujuei-oi/. 

and inventive minds to let in light on a dark subject." toCto ii ianv ai-ajiinjcris, etc. Cratyl., 399 C, fidi'oi' ruv Bripiav 

They belong to the " theories which have arisen in in- opSois 6 avOptairiK avdpuTro^ uvop.dtrSij, dvadpSn' 4 oirwirei'. 

genious minds from an imperfect conception of the pro- phmdo, 75 B, on vivra. ra ei- Taii aifffl^Veaii' eneiVov re hpiytrai, 

cesses of abstraction and generalization." But it is not ^5 o Io-tii/ Xirov, etc. 
really thinkable that the author of the Sophist, Politicus, 

and Phoedrus (249 B) did not "understand" the common- i»' Parmen., 132, ydw«« oUiH,; St-to! f, ov/c o^tos; 

sense explanation of the universal through abstraction and • ■ • ''™ o*" ''«« •""'" """ " -"ovM^-o^ Iv elvai, ie. oy to 
generalization. He rejected it, on the contrary, precisely «"" «"•' '''«'"'; Lutoslawski, p. 403, misquotes and mis- 
beoause he foresaw that, if consistently carried out and interprets this passage. Peofessoe Ritchie, Plato, pp. 
accepted as the final account of the matter, it leads 91. "2, 113, recognizes that it is conclusive against con- 
straight to Mill's ultimate phUosophy, which he would not ceptualism. Cf. Zblleb, p. 668. The further objection 
have on any terms. ^^^^ '^ ^^^ ideas are thoughts and things partake of them, 
.„' , , , . , , . .,, things must think, is generally treated as a verbal equivo- 

^^Protag., 330 C,, S«acoav., ^pay^. rj «rn. , ovS,v ^^y^^_ Euthydem., 287DE. But, for the underlying 

^pay^a; PAosdo, 65D, «a;... n eira. ««ato. avTo , ovj..; 76E, metaphysical problem, see my discussion of Aristotle de 

77 A, <aK6y T. «a. ayaOov, 100 B, /caAo. avrh KaB^ avro k^ ayagoy. ^^ 429 6 26 in 4. J. P., Vol. XXH, pp. 161 ff. 
Theoetet^ 157 D, «l cot apcVxet to fir} n elvat aWa yiyv^aSai. act 

ayaehv Kcu KoJ^ov, Cratyl., 4^B, ei 66 , . . . eort fie to Kokov, ^^^Cf. the characterization of positivism or phenom- 

eo-Ti 6a Tb aya96v. Sophist, 247 A-B, t6 yt Svvarov t^i irapa- enalism in .Rep., 516 CD, KaOopuvri to. irapiovTo. koCi fxinjfio- 

yiyvtaBai KM. attoyCyvetfOa.t. iravrus tlvai Tt ^ijtrova'iv .... x'cvoi'Ti ^aAioraf otra re irpdrcpa avriav Kal varepa eltoOei cat afia 

ouoTj? ovv fiiKatoavrf^c, etC. Philcb.y 55 B, ir«s ovk aXoyov eo-rt vopeveaOai. Cf, also PACBdO, 96 B C ; Gorff., 501 A B. 

30 The Unity of Plato's Thought 

point is not explicitly made by Plato, a concept of the mind, even apart from objective 
reference, either is or is not an entity of another than the natural or sensuous order. 
If it is, we are driven back upon Platonism. For, though the Platonic ideas are more 
than thoughts if thoughts are only decaying sense, thoughts, if radically different 
from sensations, become entities that may assume the rSle of Platonic ideas, as they 
do in the ultimate philosophy of Aristotle, and in the interpretation of those Pla- 
tonists, ancient and modern, who conceive the ideas as thoughts of God. This is 
not Plato's doctrine, but only a plausible development of it by those who cannot 
acquiesce in his wise renunciation of systematic dogmatism."" In these matters Plato 
affirms no more than is necessary for his fixed faiths and purposes."" The objective 
reality in some sense of ideas (but no more) was so necessary. That it was a hard 
saying is as well known to him as it is to his critics. '°'' And he has anticipated their 
objections. But this doctrine, or something equally and similarly paradoxical, was 
and is the sole alternative to a philosophy which he and the majority of his modern 
critics cannot and will not accept. The burden of proof rests heavily, then, on those 
who affirm that at any time he did or could abandon or seriously modify it. A survey 
of the dialogues discovers no evidence in support of such a contention. 

For this purpose the dialogues fall into three (or four) groups: (1) Those that 
are supposed to precede the doctrine ; or (2) to lead up to it ; (3) those in which it is 
most specifically affirmed or mythically embellished; (4) those in which it is criticised 
or, as some say, abandoned or modified. In the case of the first and fourth group the 
argument is often made to turn upon the meaning to be assigned to elBo';, IBea, and 
other terms elsewhere distinctly appropriated to the transcendental idea. We are 
repeatedly warned that the mere use of the words et8o9 and IBea is no evidence of the 
transcendental doctrine. This is obvious ; but it is equally true that the possibility of 
taking these words in a conceptual sense raises no presumption that they must be taken 
in that sense exclusively and that the doctrine was absent from Plato's mind at the 
time. Such an assumption is made by modern critics in the interest of theories of 
development, or to free as many dialogues as possible from the dis'tasteful paradox. 
But Plato was always at liberty to use the terminology of the ideas conceptually for 
the practical logical uses of definition and classification — even in the transcendental 
Phcedrus."' All Platonic ideas are concepts. It does not follow that they are ever 
in Plato's intention no more than concepts. And, in any case, the absence of the 
theory from any given dialogue proves no more than does the virtual absence from the 
Laws of all metaphysics, including the " later " theory of ideas. 

190 Cf. infra. Part II, Philebus. t^^^ t^^ p^^^ 3^^ Republic; LtJTOStAWSKl (pp. S«, 341), 

191 Jfeno, 86 B, «cal ri ^e^ Yt «^a ovK S.y Irani iirif, Tou tij^t it must be later, because, if we interpret rightly, we 
\dyo« Sutrxypiaai^rii', etc. " soon get quit of the riddle of self-existing ideas " and per- 

i92Kep., 532D, 476A; Parmen., 135BCj Phileb., 15AB; ceive that " iSea and e'Sos are used in a meaning -which is 

Tim., 51 C D ; infra, p. 36. identical with the idea as conceived by Kant, a necessary 

193 237 C, 249 B, 263 E. Cy. also the loose popular use of concept of reason." Of course, Kant's ideas of reason are 

«I5o5 and iSe'a 237D, 238A, 253 CD. NATOfip, Hermes, Vol. misapplied here and all Lutoslawski means is " Begriff," 

XXX.V, p. 409, infers that the Phcedrus, " must" be earlier "concept," 

Paul Shoeey 31 

Premising thus much, we turn to the first group. In the Apology, Crito, Laches, 
Lysis, Charmides, Menexenus, first four books of the Eepublic,^^ Proiagoras,^^' and 
(some affirm) the Euthyphro, Gorgias, and Euthydemus there is no distinct mention 
of the (Platonic) ideas. There was no occasion for it in the Apology, Crito, and 
Menexenus, and little, if any, in the others. The relation of the Lysis, Charmides, 
and Laches to Plato's mature ethical theories and the subtlety of the Charmides and 
Lysis^"^ make it improbable that they antedate the main tenet of his philosophy. This 
is still more obvious in the case of the Menexenus (387 (?), ast. 40).'" The realistic 
language used of the definition in the Euihyphro must be presumed to imply what a 
similar terminology does elsewhere."" The joke about irapovaia in the Euthydemus is 
a distinct and familiar allusion to the Platonic idea of beauty."' Had Plato omitted 
that jest, the absence of the doctrine would prove no more than it does in the case of 
the Protagoras. 

More interesting than this balancing of probabilities is the evidence presented by 
the Gorgias. This magnificent composition may or may not be earlier than the Meno, 
Phcedo, Euthydemus, and Cratylus. It is certainly not appreciably less mature. It 
distinguishes and classifies "ideas" in the manner rather of the "later" dialogues,""" 
and although it contains no explicit and obvious mention of the transcendental idea,""' 
the doctrine is clearly suggested for all readers who look below the surface. It is 
worth while to dwell upon the point. In the Cratylus, 389 C, employing the termi- 
nology of the ideas in the manner of Republic, 596 A B, 597 B,™'' Socrates says that 
the workman who makes a tool puts into the material, the iron, the idea of the tool 
that exists in nature.*" Similarly in Republic, 500 D, the philosopher statesman puts 

19* 402 C and 437, 438, presumably imply the ideas, but eto.,seemy remarksin^. J. P.,Vol.IX, No.3, p.287.) More- 
could be taken merely of concepts, classes or species. Not over, Plato never affirmed the presence absolutely of the 
so 585 in Book IX. Pfleiderer therefore, in order to elimi- idea with or in the particular (Parmen., 131 A B; Phileb., 
nate the ideas from Books VIII and IX, pronounces 580 B- 15B), but only its presence or communication somehow. 
588 A a later addition. The rt of KaXAo? Tt expresses this and Socrates's embarrass- 
es But Cf. 330C, i? fiiicaiiKrwTi TrpSvitia Ti ecTTi.- 5 ova'ei- ment very well. Cf. Ph(Zdo, lOOD, elre wapovjia tire Ko^v»,via. 
wpiyii.a; 349 B, t) Uitrrf tSip ovoiLirav tovtioi/ viroiceiTai Ti! ISios "^ ""U ^'I «»' '•'""5 irpoffyero^eVi,. So Symp., 211 B, ixtTtxovTO. 
oiro-ia Kol irpay/na ixP" ««"ToO SvvaiJ.i.v inaaTov ; 330 E, ovtt) f, TpoTrov Tii-i ToiouTOi', etc. 

oirtiini!. 200 JowETT, Vol. IV, p. 436 : " The same love of divisions 

19S Cf. supra, u. 108. is apparent in the Gorgias." Cf. 454E, 455 A, in manner of 

mwilamowitzhas somewhere denied the Platonic *^« f'?'^?!? • C/;«*'«5. It could be pUusibly argued that 

authorship of the Menexenus, but he may have a "pec- the definition of rhetoric .<,K.t..„, ^op.ov e.S»,W (463D) as 

.„. T-»ui.4.jv4 I, j« . explained m 464, 465, IS a fuller and more explicit statement 

cam " in reserve. Life is short to debate such paradoxes j ^ .„,..■ „-,, r> j.,n,-u:\n,A .. 41, 

, . .. iu J.- n J. , I,- if 4.i,» „«■• t of the doctrine of Po!t*ic««, 291 B and 303 E-304 A, as to the 

but if any athetizer will stake his reputation on the point, ,._ ,, , ,. .^. . , . ,, . . . t- • .1 

... ,, , difficulty of distinguishing the statesman from his imita- 

tors and the true relation of piTopeia to 6«a<rTiic7) and 
198 5 D, TavT6v .... avrh auT<p, etc. ; 6 D E, rSiv iroWuiV BmtlXlkv. 

oiriiov [cf. PftcEdo, 78 D, Ti Se Tuv iToAXii' (cuAii'; iJep., 596 A, „„, t. ^ ^ i... t> . o, • joot, - ->, 

• . , . o N .,v. i . . tt * 201 But cf. 474 D, an-oSAeTTwi' ; 488 D, ra riav itoKaihv fo/xiua, 

elfios .... er ... . irept CKOiTTo Ta iroAAa) — auTO to eI6o5 w, ... _ . «t. . - . , .. . . l ^ 

/r... _, .nr.-r. ~ , ' ..r nn rf " Tr With 2?ep., 479 D, Ttt T(OV ITOAAwi' ITOAAa VO^l/ia, OtC. ; GOTg., 

etc. (Pftoedo, lOOD, T<j>itaAii);-JfejM>, 72C, ei/ye Ti tWo! . . . . , , . ,,, . 

fit' &), airop\eiTbiv .... TrapafieiyjLtaTt, 

497 E, irapovtria . . , , oU av /caAAoc irapy. 

199 301 A. It is not the word irapeaTi that proves this, a- ' 7 i ■> 
but the entire context cTcpa avToO ye ToO KoAoi), etc. Ltjtos- 

LAW8KI (p. 212) affirms that Plato "would have said later 203 On this passage as the chief Platonic source of the 

iripeo-Ti TO xaAAo; (avTb ica9' oiiTd)." He never did say, nor Aristotelian doctrine of matter and form see my remarks 

could he have said, anything of the kind. HapeaTi .... in A. J. P., Vol. XXII, No. 2, p. 158. Campbell, overlooking 

airb icafl' auTo he would have felt as a contradiction in this passage, finds in Polit., 288 D, the earliest approach to 

terms. (On the correct and incorrect use of avra wafl' ai/ra., the distinction of matter and form. 

32 The Unity of Plato's Thought 

into the plastic stuff of human nature the forms or ideas of justice and temperance 
which he contemplates as existing in the transcendental world (e'«et), and so becomes 
an artisan of political and popular virtue.^"* Expressed in slightly different imagery, 
this is the function of the statesman in the Politicus, 309 C (c/. 308 D). He is to 
implant in those rightly prepared by education, fixed, true opinions concerning the 
honorable, the just, and the good.'"'^ The thought and the imagery belong to Plato's 
permanent stock. We find them in the Gorgias, 503 E-504 D.^°° Here, too, Plato 
conceives the true teacher, artist, or statesman as contemplating ideas or forms, which 
he strives to embody in the material with which he works, even as the Demiurgus of 
the Timceus stamps the ideas upon the matter of generation. 

The origin, first suggestion, exposition, or proof of the theory of ideas is variously 
sought by different critics in the Meno, the Cratylus, the Thecetetus, or even in the 
Phcedrus, Parmenides, and Symposium. Obviously Plato could at any time argue 
indirectly in support of the ideas as necessary postulates of ontology and epistemology. 
Our chief concern is with the hypothesis that the exposition of some particular dia- 
logue marks a date in the development of his own thought. The doctrine of remin- 
iscence is introduced in the Meno to meet an eristic use of a puzzle allied to the 
psychological problem of " recognition." ^^ How, if we do not already know, shall 
we recognize a truth or a definition when we have found it?™' Socrates replies that 
the soul has seen all things in its voyagings through eternity, and that all our learning 
here is but recollection.^"' This theory is confirmed in the case of mathematical ideas 
by Socrates's success in eliciting by prudent questions a demonstration of the Pythag- 
orean proposition from Meno's ignorant slave."" The Phcedo distinctly refers to this 
argument as a proof of the reality of ideas,^" and the myth in the Phcedrus describes 
the ante-natal vision of the pure, colorless, formless, essences of true being. '^'^ It fol- 
lows that, though the ideas are not there explicitly mentioned, the reminiscence spoken 
of in the Meno must refer to them."' But it is extremely improbable that this repre- 
sents Plato's first apprehension of the doctrine. Psychologically and historically the 
origin of the theory is to be looked for in the hypostatization of the Socratic concept 
and the reaction against Heracliteanism."* Its association with Pythagoreanism and 

204 a elect opif /leAeT^irai ct9 avBptoiriav rj&rj .... TLdei'ai 207 ji/e/io^ 80 D ff . Cf. my dissertation i)e Platonis ideOr 
. , . , iritLLovpyov .... <Tit>ti>potrvyj}v t« koX SiKaLotrvvTii, Cf. rum doctrina, pp. 15 ff. 

501 B, TTVKva ay eKarepwcre aTTOjSAeTrotej' .... koX Trpby €Ktlvo a5 onn - ^ - - • - - v ' . « j ti ■ 

".'.,„, , - ^., T, 7-i onnt^ - • - zlWouTe iriTtiv ouTc anopciv avev rrpoAi)i/(fco5, Sext. Empir. 

h ev Toiff arOpwn-ois e/Airotoicr, Cf. FoUt.y aJd 1>, touto auTO Mnth 1 ^7 


205 This does not refer exclusively to the higher ednca- '"' " I^'^i^'s pent dire comme le Dieu do Pascal : ' Tu 
tinn as Zeller affirms °® ™® chercherais pas, si tu ne m'avaisdfij^ trouvfi.'" — 

' ; ....... FouiLLfiE. C/. PoZtt., 278 D; Tim., 41 E,Ti,^To07raiT65*vViK 

206 aT^o^AET^(l>f Trpos Tt .... ottws av elfios ri aurw trxrt toOto tfietf e. Cf. infra p 43 
o epya^eTOLi. This is applied first to the body, then to the 

soul. The Tafis and icd<r^os of the soul is StKaLoavvri and odii. 

ffia^po(rvv7} irpb? ravra ^Kiiriav^ etc. The ptjrwp aya-Soi 211 73 A. 

•cdl TevKuco! here = the true itoXitikos. And we may note in v><nntr oinn - t. • • . • • 

. .1 . ..!_ ^ - (I 1 J „ it i V i ^^^ 647 II., 249 C, TOUTO 0« effTti' arauFTjo-ts €Keii'wi', etc. 

passing that the Goro«M already " recognizes that rheto- ' ' 

ric might be an art. The popular rhetoric is none because 213 The realistic terminology of the definition would 

it ignores the ideas (1) as ethical ideals (Gorgias), (2) justify the same inference. Qf. 74,75. 

as the basis of scientific dialectic (.Phoedrut). 2U Cf. supra, p. 28. 

Paul Shobey 33 

the ante-natal life of the soul is mythical embellishment ; and its application to the 
problem of the a priori element in human knowledge is a secondary confirmation of 
its truth."' Nevertheless the Meno, which John Stuart Mill pronounces " a little gem," 
is admirably adapted to serve as an introduction to the Platonic philosophy. It exem- 
plifies in brief compass the Socratic method and the logic of the definition in termi- 
nology that suggests the ideas, touches on higher things in the theory of recollection 
and the problem of a priori knowledge, and clearly resumes the dramatic, ethical, and 
political puazles that prepare for the teaching of the Republic. Socrates's mention of 
the ideas at the close of the Cratylus as something of which he dreams as an alterna- 
tive to Heracliteanism is taken by some critics to indicate that we have here an intro- 
duction to or a first presentment of the doctrine/'' They overlook two considerations : 
(1) the theory is taken for granted at the beginning of the dialogue, as we have already 
seen;"' (2) there are no traces of immaturity in the thought of the Cratylus. The 
polemic against the flowing philosophers and the forms of eristic associated with them 
is, in a jesting form, as sharp, and the apprehension of the real issues as distinct as 
it is in the Thecetetus and Sophist."' 

Some scholars look upon the Thecetetus as a propaedeutic introduction to the 
ideas,"' while others take it as marking the transition to the later theory. Strictly 
speaking, neither view can be correct, since, though the ideas are not often or very 
explicitly mentioned, there is enough to show the presence of the doctrine in its normal 
form. The ayaOov and kuXov, claimed for being as against becoming in 157 D, is 
almost technical for the affirmation of the ideas. '^° The TrapaSeiy/iara of 176 E can 
hardly refer to anything else. And the close parallel between 186 A B and Republic, 
523, 524, admits no other interpretation. Among the vor)Td which the soul grasps by 

215 Pbopessor Ritchie's suggestion {Plato^ pp. 86, 87) tinctly implies that tto-vtil pel includes qualitative change, 
that the Platonic idea is a generalization of the Pythag- Cf. 439 D, on toioutoc, 440 A, aXAo Kal aWotov yiyvotro .... 
orean treatment of mathematics is unsupported by evi- inroloy ye tl etrrtv. 439 E, tniSev i^iarafievov t^s avrov LSea^ {cf. 
dence. See, however, Zelleb, pp. 654-6, for suggestions Tim., 50 B, and Sep., 380 D). C/. the whole context of the 
of other pre-Socratic influences on the theory. argument and the useofun-e^epxcatiCrafyl.i^D; Thecetet. 

216 So once Susemihl in his Genetieche Entwickelung, 182 D. In fact, the association of motion and qualitative 
Vol. I, p. 161. LCTOSLAWSKI, pp. 224, 225, thinks the ideas change was always a commonplace with Plato. Both " be- 
are not formulated even here, but only a something which 'o™ " an* " a'ter " the The<Btetus ^eraPo\rj and pely, etc., are 
in later dialogues proves to be the ideas 1 The terminology "sed freely in both meanings. Cf. Bepub., 380 E ff., which 
iscomplete-eWos, avrb o tVn, t6 <^>;<r«i, irot ^AeTrco;. 389, tl ii ^^°^^ refutes L.'s "discovery." oi.oS- v;rb y.iv iWov Ti 
.... eVri Si TO KoKoy, i<m Si rb ayaBSp, eVri «« eV eVaarov riv •C"'™ '"X""^" 5«'"° iAAoioflrai re Kal Kiytlra,. The fact that 
o..TOr (440 B). All these phrases might conceivably be used "^e ThetBtetus is slightly more explicit in formal olassiflca- 
of notions, conceptual ideas. But this proves too much. ^^°^ P™™^ nothing. The whole argument of the Cratylus 
For, according to L., it holds of all dialogues except the Passage hinges on the distinction precisely as does the argu- 
Symposium, Pftcedo, and parts of the Republic, and he is m™' <>' the Thecetetus. It appears explicitly again only 
not quite sure of them. His real object is to eliminate the *" *•>« Parmenides, and not in the " late " PhiUbv^ and 
self-existent idea altogether. Tinuzus. It is not included in the ten kinds of motion in 

„^ J 1 ■ I » J • non ■ the Laws, 893, 894, and L. finds it only by implication in 

217 Cf. supra, p. 31. The doctrine of CratyL, 389, is ^^ .^, .^^_ .^^ ^.^^ ^. ^j, ^^^.^^^ .^^^^^ iKKoui^T,, which 
furthermore identical with that of Bepub., 596 A fl. .^ ^^ ^^^^ ^^pjj^.j ^.j^^^ ^^^ Cratylus or Bepublic. 

218 386, 439, 440. On the ini ov and i/(evS^5 Sofa fallacy, 

429 a., cf. infra, p. 53. On the p^^c, cf. 411 B C with "''W. J. Aleiandeb in Stvdies Dedicated to Gilder- 

Phcedo, 90C;PAiie6.,43A. Lutoslawski affirms (pp. 366, sJeewe, p. 179, thinks its teaching to be : knowledge is of the 

367) that the subdivision of «..n,Tti into *opa and aWo;«.a« ideas, error arises from imperfect i,'i^vr,a,,. 
is a new and important discovery of the Thecetetus, 181, C. ^"Supra, u. 185. 

He fails to note that the argument of Cratylus, 339, 340, dis- 

34 The Unity of Plato's Thought 

herself,^^' and whose essence is apprehended through their relation of opposition/^'' 
are mentioned, after ova-ia, the ofioiov and avofioiov, the ravrov and the erepov of the 
Sophist. But also, as in the Parmenides, the ethical ideas, kuXov, aiaxpov, ayaOov, 
and KaKov ;^^' and lastly, as in the Republic, the qualities of sense, aKXrjpov and pLoXaicov!^^'' 
The actual sensation of these opposites comes of course through sense. But the oixrla 
and the 6 ti ia-rov, as in the Republic, is apprehended by the mind as an idea. There is 
no argument for holding these ideas to be mere concepts that would not prove the same 
for the Republic, which of course is impossible. '^^^ This point established, we may 
concede that the Thecetetus may be, not an introduction to the ideas, but an indirect 
argument in support of the familiar doctrine. The polemic against Heraclitus is 
always that.^^* And, though Plato himself may not be aware of it, the statement that 
the syllable is not the sum of its elements, but fiia Ihea a/j.epia-TO';, embodies the prin- 
ciple and justification of a realistic logic. ^' The conceptual whole is not the sum of 
its parts, but a new entity and unity .^^* 

What has been said of the Thecetetus applies to Zeller's theory ^^' that the second 
part of the Parmenides is an indirect argument for the ideas. That this is not the 
main purpose of the Parmenides will appear in the sequel. And Zeller was mistaken 
in stating that only relative contradictions followed from the being of the one, while 
absolute contradictions resulted from its not being. But the Platonic idea is always 
suggested by the antithesis of the one and the many. And in the eighth hypothesis, 
164 B ff., the "one" and "others" are no longer treated with dialectical impartiality, 
but there is a hint that the one may be regarded as the symbol of the idea. Symme- 
try leads us to expect the argument that, if the one is not (relative /u.^ 6v), other things 
both are and are not all contradictory predicates. Instead of "are" we find " appear" 
or " seem." Other things are indefinite bulks that break up under inspection and 
only seem to partake of unity and other predicates that derive from unity. These 
07/cot certainly suggest the world of matter uninformed by ideas, the " being" of the 
materialists which the friends of ideas in the Sophist call " becofning" and break up 
into little bits.^^" And the statement that, as they cannot be other than the (non- 
existent) one, they are the other of one another, reminds us of aX\i]\oK .... aw- 

221 avrri ^ i^ux^l86B. C/. 187A; Phmdo, 65 C; Bep,, 524 paradox, but passes on to show how mathematics leads 
B C, 526 B. the mind to the apprehension of abstract and ideal unity. 

222 tV ivavTionrroL irpbt iW^Kio. Of. Sep., 524 D, « liiv eii Fhilebus, 14 D £E., is concerned with logical method ; Bep., 
Tijv ala-eriiTi.i' i)t.a Tols ivavrioit iavrolt i^mmei. Mr. Henry ^23-6, with psychology and education. But the thought of 
Jackson and others confound this special use of Trpbt aXX.)A<i the BepuftJtc is not less mature, and is, indeed, repeated in 
with tA irpdi Ti, relative terms generaUy, by the aid of Par- Phileb., 56 E =Bep., 525 D E. 

men., 133 C. The rfteoetettM passage is the source of Her- 226 ovSky eXvai h/ airo KaS'' avTo, etc. 157 A is the diametri- 

modorus's distinction of jrpbs irepa into npo^ evavria. and irpd? cal opposite of the ideas — elvai tl KaKbv avrh Ka0' aiiTo, 

TI, which Zbllee (p. 706) says is not found in Plato. Phcedo, lOOB. 

223 130 B after o^oidn)!. 227 205 C, 203 E. 

224 186 B with iSep. 524A. ^^^Cf. Parmen.,151J), ovk apa rmv iroWiiiV oiiSe navTtav TO 

225THOMPSONon Jl/eno, 74D,says that problems which „ "^ , "^^ L. r, ii ^ tt-ittt >t « -Ic >-<"«»'. 

„,.,,,,_ ,. , J iu u c See also J^. J. P., Vol. XXII, No. 2, p. 158. 

in Phileb. ,UT3, are Sc6rifj.tvii.eya are made the bases of a ' i "• i f ""• 

dialectical course in Bep., 523-6. This is a misapprehen- 229Set forth in his Platonic Studies and the earlier edi- 

sion. The Bepublic mentions (525 A) that the same object ''""s of his History, but now virtually withdrawn. 

is perceived as one and many. It does not sport with the 230Spp7i., 246 B.C. 

Paul Shobet 35 

SeSeffdai in the theory of pure relativity in Thecetetus, 160 B. Similar hints occur in 
the fourth hypothesis, 157 B, which deals with aWa on the supposition that the one 
is.™ The main conclusion that aXXa, then, admit all contradictory predicates is indi- 
cated very briefly (159 A). What is emphasized is the fact that aXKa per se are 
ifKridr] . . . . iv oh to ev ovk evi, that they are airetpa (cf. Phileb.) ; that it is the one 
which introduces Tre/jo? Trpo? aXXrjXa ; and that, having parts, these parts must relate 
to iud<; Tcvbi lBea<s Koi iv6<; tivo<s, 6 KaXov/iev 6\ov.^ While the main object of the 
Parmenides, then, is to illustrate the communion of ideas and the doctrine of relative 
01/ and fii] $v set forth in the Sophist, there is a suggestion of polemic here and there 
directed against the infinite and indefinite world without unity of the materialists, 
relativists, and deniers of the ideas. But obviously the first origin and exposition of 
the ideas is not to be sought in a work that deals with problems and diflBculties arising 
from the doctrine.^'' 

The Phcedo, Phcedrus, Republic, and Symposium, the dialogues that are fullest 
in explicit affirmation or mythical embellishment of the transcendental idea, need not 
here detain us long. In his exaltation of pure thought and the dialectical method 
Plato clothes the ideas in all the contradictory attributes of a sensuous, aesthetic type, 
an ethical ideal, and a metaphysical noumenon. He is perfectly aware of this, and the 
inconsistency is common to all philosophies of the absolute.^" In the Phcedrus as 
elsewhere he warns us not to take the myth too seriously .^^ In the Phcedo he 
describes the doctrine as familiar,^'* and reminds us that he does not insist upon the 
precise terminology, but only on the central fact.^" In the Republic every termi- 
nology is employed from the most naitve to the most severely logical or the most 
transcendental.'^ Despite these facts, attempts have been made to extract evidences 
of contradiction or development from the varying imagery and terminology of these 
dialogues. The unity of the Republic has been broken up and its books variously 
dated according to the absence of the theory, or its presence in an "earlier" or 
"later" form. It has even been gravely argued in defiance of all psychological and 
historical probability that the Symposium, which in consonance with its theme men- 
tions the idea of beauty only, represents a stage of development in which the Platonic 

231 Belative ov admitting Koivrnvia. Socratic ethical concepts, not Platonic ideas, is refuted by 

232 The<etet. 203 204. ^^® context (19 Totavnj ovo-i'a .... ava^epuy TO. £V Tats aitrd-^' 

233 Cf. infra. Part II. 

treat, etc.). The suggestion that the reference is to con- 
versations abandons the whole case, unless they are limited 
234JowETT's common-sense and literary tact have an- to jj^e interval between the Meno and the Phcedo.' The 
swered literal-minded objectors once for all: "When the simple truth is that Plato may at ahy time refer to any 
charioteers and their steeds stand upon the dome of heaven pa^j ^f ^g permanent beliefs as familiar doctrine. On the 
they behold the intangible, invisible essences which are theory of development, to what discussions is reference 
not objects of sight. This is because the force of language ^^de in Crito, 46 D, and 49 A B 7 To the Gorgias and Be- 
can no further go."— Vol. I, p. 412. public 1 1 Where has Plato often said that to tA airoS 

235 265 C, tA jiiei' aAAa Tu oiTt iroiiSif ireiraio-Soi. irparTeii' is Sucaio<rvVi| ? (Sep., 433A). Where has Glaucon 

heard ovk oKiyajci^ that the idea of good is the tJt^yurrov 
li.ieriii.a 7 (.Bep., 504 E). 

236 Those who think that the ideas have been men- 
tioned in only one preceding dialogue, as the Meno or Syni' 
posium, are much exercised by the Sa/xa Ae-yeiK of 72 E, the ''^' '^ ^• 

i epv\ovpi€v iei of 76 D, and the voKvBfvK-itra of 100 B. Lnios- 238 596, 597, 5g5, 534, 532, 514-17, 505-11, 500 B-501 B, 490 B, 

LAWSEi's statement (p. 292) that these terms may refer to 485 B, 476-80. 

36 The Unity op Plato's Thought 

philosophy contained but one transcendental idea, as if the problems of psychology 
and ontology which the theory of ideas sought to meet or evade could have been in 
any wise advanced by the hypostatization of one concept ! We have glanced at such 
methods of reasoning already, and shall meet them again. At present we pass on to 
the hypothesis that the Parmenides contains a criticism of the ideas which leads to 
the abandonment or transformation of the theory in the fourth and latest group of 
dialogues. This hypothesis rests on the assumption that the criticism of the Parmen- 
ides is new, that Plato was bound either to answer it or give up the ideas, and that, 
as a matter of fact, the transcendental idea is not found in the later dialogues. These 
assumptions will not bear critical examination. 

The objections brought forth against the ideas in the Parmenides are obvious 
enough, and, as Jowett says, are imanswerable by anybody who separates the phe- 
nomenal from the real. How can we bring the absolute into intelligible relation with 
the relative? How can the absolute (" the Gods") take cognizance of us or we appre- 
hend what is adapted to their thought?^" How can we without self-contradiction 
apply to it unity or plurality, or any other predicate of human knowledge?''"' More 
specifically, if the ideas are transcendental unities, how can we predicate multiplicity 
or parts of them as wo must to connect them with one another and with phenomena ?"' 
How shall we interpret the figurative expressions that the ideas are present in things, or 
that things participate in or imitate the ideas ?^*^ If the idea is the postulated corre- 
late of every idem in multis, why should we not assume an idea to explain the likeness 
of the idea and the particular, and so on in infinite regression ?"^ To what extent the 
form of these objections is due to contemporary critics, or the misunderstanding of 
students, or the precocity of Aristotle, is an unprofitable inquiry. Their substance is 
in the Republic, not to speak of the Phcedo, the Euthydemus, the Timceus, and 
Philebus.'** Their presentation in the Parmenides, then, does not mark a crisis in 
Plato's thought calling for a review of his chief article of philosophic faith. Plato 
does not and cannot answer them, but he evidently does not take them very seriously,"' 
though he admits that it would require a marvelous man to sift and analyze them all."* 
They arise from the limitations of our finite minds."' Here as in the Philebus he 
bids us disregard them, and proceed on the assumption of ideas to find the one idea 

239 Par-men.^ 134, Sophist, only because pedants were obstructing the way of 

2'OSopft., 2M, 245; Parmen., 142 A; Tim., 37 E, 38 A. 1°^« ^^ denying it. Similarly the rpiVos ivifiajTOi is dis- 

„., „ ,„ _, ., . i,Tj tinctly implied in Republic, 597 C, and Tim., 31 A, as the 

24iPormen.,131; PAjie6.,15B. j-^ u. . • • ■ ■ ^ , . . 

dilhculty of giving a precise meaning to irupovo-ia is in 

212 Parmcn., 131 A, 132 D. 2«132A,132E. E«tM,dem«8, 301 A, and PAcedo, 100 D. 

S"Bep., 476 A, airt ii.iv iv eKovrov elvai, tr 8* rwv irpafeuK Ub pMlgb., 15DE. In Sophist 251 BC, the reference is 

«al (Tio^aTu^ Kai iXA^Xa)^ KOKvayia wo\ki c/«ure<r9ai ticcwTTo^. ^^ j^e One and many in things, but the application to the 

Cf. Phileb., 15 B; Parm.en., 144 B. Some ignore this pas- communion of ideas immediately follows. 

sage. Others wantonly emend it, as Badham, who reads 

oAAj aAXoji', and Biwatee, who reads iAA' ofAi^v {Journal "epormen., 135AB, 

0/ PTis!., Vol. V, p, 122), ElTCHTB (Pfato, 96) takes it in a 2*7 Tim., 52BC, 34C; Phileb., 151), rav Xoyui- . . . 

Pickwickian sense in order to avoid "anticipating the Tlie 5opftt«f does not really contradict Tim., 

Sophist." PpiiEiDEEEB uses it to prove that the fifth book of 38 A B, Absolutely ov and (iij ov remain a mystery (251 A, 

the iJepubiic is later than the tenth. Anything rather than 251 D, 254 C). The Sopfttsf merely flies the practically 

admit the obvious fact that Plato always recognized the necessary conventions of logical discourse about them — 

"communion" of ideas, and argued it at length in the toc Adyoi', «»• toIs wop' ^^iv Aoyois, etc., 251 A, 251 D, 

Paul Shobey 37 

and enumerate all its species."' The hypothesis must be judged by its total con- 

The text of the Parmenides does not bear out the assertion that the objections 
apply to any special form of the theory or can be met by a change of terminology. 
The suggestion that there may be some classes of concepts to which no idea corre- 
sponds is repudiated for good Platonic reasons. ^^° The interpretation that the ideas 
are to be henceforth merely concepts is distinctly rejected, was a priori impossible for 
Plato, and is refuted by the positive affirmation of their objectivity in the Timceus^^ 
Socrates's explanation that the ideas are irapaBeiyfiara, patterns of which phenomena 
are likenesses, is nothing new. The terminology of pattern, copy, and artist looking 
off to his model is familiar throughout the "early" dialogues, whether used of the 
definition or the idea. There is no hint in the corresponding passages of the Philebus 
that such a variation of terminology could in any way affect the problem. It is not 
proposed in the Parmenides as a new doctrine, but merely as a different metaphor to 
evade the difficulty foimd in the literal interpretation of nerexeiv — it is a mere gloss 
upon the meaning of fierexeiv. But equally formidable difficulties confront this way 
of putting it.^^'' And there is no systematic change of terminology in the "later" 
dialogues, which, like the earlier, employ in a purely natural and non-technical way the 
various synonyms and metaphors which Plato used to express the inexpressible.^^' 

The challenge to find the ideas in dialogues "later" than the Parmem'des is easily 
met. Nothing can be more explicit than the Timceus.'^* The alternative is distinctly 
proposed: are the objects of sense the only realities and is the supposition of ideas 
mere talk ? "° And it is affirmed that their reality is as certain as the distinction between 
opinion and science. They are voov/ieva and exist Ka9' avrd.'^^ There is no hint that 

2<8135BC, Pftiieft., 16D. (y. PAoedr., 270 D, eav {i jrXciu lar in the idea. The o|u°"^f''"'<' are no more separable as 

€15?; £XH TavTa aptdiitjiratievovi. Laws, 894 A, iv elSeo-t Kafislv an intermediate stage than are ra elinovTa koI e^iovTo. rSiv 

fL€T^ ap^Bfiov, ovTtitv aei fiiti-^iJLaTa of Timceus, 50 C. In both cases we have 

2*9Parme7i. 136 • Phoedo 101 D. ""^^ '^^ '*^®® ^°*^ *^® particular and the metaphorical 

«,».„.-» r, n' ..»...». a .• i • -J T> i Bxpresslon of their relatloH. 

250 130 D. See Zeller, 700, 701, for lists of ideas. But, 
as we have seen, to admit that there is any conceptual "' See my note iaA.J.P., Vol. X, No. 1, p. 66. Zblleb, 

unity not referable to an idea is to make the theory a mere Sitzunasber. d. Berl. Akad., 1887, No. 13. 
play of fancy, and deprive it of all psychological and onto- 251 51, 52. 

logical meaning. 255 51 C, rb Se oiSiv ap' V t^'I" ^"Yos. For the impossi- 

2ii 51 C. Cf. supra, n. 188. bility of taking Advos as " Socratic concept " see my note in 

252 The TpiTos ii-SpcoTTos is repeated in 132 DE. Other -4. J. P., Vol. X, p. 65. 
difficulties follow, and the final summing up, 135 A, is 256 Me. Abcheb-Hikd's attempt (Jour, of Phil,, Vol. 
couched in the most general terminology : ei eitriv atrai at XXXV, pp. 49 ff.) to "circumvent" this passage is based on a 
ifieai rS>v ofTuf Kai bpietrai Tts avTo Tt eKeurrov eiSo^, There is misinterpretation of 39 E. Since an idea of fire is not men- 
no suggestion that a. new form or terminology makes any tioned in the exhaustive enumeration there given of the 
difference. The much misunderstood passage, 133 C D, is ideas contained in the supreme idea, an idea of fire he 
merely a special application of the general difficulty to argues, cannot be meant seriously here. But 39 E does not 
relative terms. Ideal slavery is related only to ideal speak of the " supreme idea," which is a figment of modern 
ownership, the slavery in us only to the ownership in us. Flatonists. The ivov is simply the universal of animal or 
There is no discrimination here of a class of aira. <a9' aura living thing, and as such the paradigm of the world which 
e"i«i). (Cf. A. J. P., Vol. IX, p. 287). Nor are there, as Jowett is a living thing. (.Cf. A. J. P., Vol. IX, p. 294.) It includes 
and Campbeli. affirm (Bepublic, Vol. II, p. 313, u. 1) two all subordinate voTiTa f^a. There is no reason to look for 
stages (1) 6/iot'w<rts and (2) tieQe^tt tov o/ioiu/iaros in the other ideas in it. J. Hoeowitz {Das PlaUmische voiyrhv 
descent from the ideas to the individuals. op-oMiiara and iSiov und der PhiUmische «o<rjiidt votito!, Marburg, 1900) fails 
fiiTexovTej are merely two sides of the same fact — the par- to prove his assertion that the vojirbv (aov is " die Welt- 
ticipation somehow (elre own 5^ TisavrdnflcTttt) of the particu- Idee." Mr. Archer-Hindis further arguments merely pre- 

38 The Unity op Plauo's Thought 

they are mere concepts, or thoughts of God. On the contrary, God uses them as pat- 
terns, and as elements in the creation of the soul.''" They are characterized in 
terms applicable only to pure absolute Being, and the familiar terminology is freely 
employed. ^^ Three things, Plato repeats, must have existed from all eternity: the 
pure Being of the ideas, the generated copies, and space, the medium or receptacle. ^^^ 
The attempts of modem scholars to eliminate these elements or identify them with 
other categories found in other dialogues contradict Plato's explicit statements. We 
are often told that space is the OuTepov or fit} ov.''^ For this there is not a scintilla 
of evidence.^" Plato even says of space: TavTov avrrjv ael irpoaprjTeov (50 B), and calls 
it a TpCrov av yevoi 6v ro tjj? ■)(mpa'i aeC. The "same" and the " other" appear in a 
wholly different connection in the creation of the soul, and are obviously the categories 
of the Sophist attributed to the soul to explain its cognition of sameness and differ- 
ence.^^ The occurrence of these categories in a dialogue that reaffirms the transcen- 
dental idea proves that to Plato's mind the two points of view were not incompatible, 
which, for the rest, is obvious enough from the Phcedrus. We must interpret the 
Sophist, Politicus, and Philebus in the light of this presumption, and treat the termi- 
nology of the ideas as prima facie evidence of the doctrine. The Republic (476) 
" already " states that the transcendental unity of the ideas is somehow compatible 
with their communion. The Sophist formulates all the concessions which a " working 
logic " must demand from all philosophies of the absolute, be it absolute relativity, 
absolute Being, or absolute Platonic ideas. Plato minimized the inevitable inconsist- 
ency, and a sound interpretation will not exaggerate it. A working logic does not 
emphasize the transcendental character of the idea. But the language of 248 A, 
247 A B, distinctly implies it.^'" The statement that hiKaioa-vvt} and (j)p6vr]cn<; are engen- 
dered in the soul [iyyiyverai) obviously does not mean that they are per se concepts 
of the mind. Nor can we infer that the ideas are mere concepts from passages in 

sent the usual objections of common-sense conceptualism — aio-^Tjo-ty, but Aoynriacp rtvi v60tfi (52 B) . But Plato's terminology 

which are not competent to anyone who himself believes in cannot bo used out of its context in this way. The (*)) ov 

any metaphysics or attributes metaphysics to Plato. problem belongs to logic. Phcenomena are intermediate be- 

257 28 A 29 A, SOB C, 35 A. Zellee, p. 665, n. 2, adds tween o^ and ^i, oy because they change, and are and are 

Phmdr., 247, which is irrelevant, and Bep., 596 A £f., where °°' '''« same predicates not because they are the offspring 

God is the maker of the ideas. Lutoslawski's argument °f ^^^f ^J"^ -"a""- I" Pl»y^«=| ^l^to was forced, how- 

from yoria,, ^eri Wyou Trep.X,™.- (27E, 29 A, pp. 474, 477) in- ®™^ reluctantly, to assign a kind of eternity to matter or 

terpreted as "included in thought " is a simple mistransla- ^P^f'- (^A Beekeley Principles, sec. 117 : " either that 

real space is God, or that there is something beside God 

'°°' _ .„ „ „. „ „, „T, . . ™r, . which is eternal, uncreated.") So far is it from being true 

258 52 A, 27 D, 28 A B, 29 B, 30 C. Cf. 39 E, o ..rrc; 37B,Ta ^^^^ ^^^^^ ^^ matter imparts ^S, 5. to pAcEnomena that, on 
<ari ravTi Ixovra «. ; 48 E jrapa6«waTo,, to which cor- j^^ contrary, Plato explicitly says that phcenomena, being 
respond 50 C, M'/x,^aTa, and 52 A, om»«<m.o. o^otov ; 31 A, the ^^^^^j j^^^^^^ ^j;^g ^^ ^^^^^^^ (^._^.^^j somehow through 
Tfiroi iyBpanoi. their existence in space. Km., 52C. 

259 CO Tj 

'82 37 ABC is plainly a psychological myth or allegory 

26» E. g., by Eitchie, p. 116. expressing the results of the analysis of the Sophist. Cf., pp. 719 £E., 733, produces none. Aristotle's also Tftecetet., 194 B. 
obscure allusions prove nothing. The identification of the mSii Koyiaiioi Si <lnxi irpos rnv ivrat ovaiar, fiv aeX Kara 

iireipov of the Philebus with Mi ou and matter breaks down. ^i^i <i<ravTus ix'^v ^uti ou<r,! oiv SuauoailvTit xal ^po- 

There remains the argument that, since in the BepjMic the „^o^e<»s .... jrorepov oparav icol aTrrti' (c/. Tim., 28 B) tXvai ^aai. 

ideas are iv and phenomena are (ierof i — SrTot and (lij oi-Tot, ^, „;^^^ jj „^„j„ dipara. 
matter must be ni or apprehended neither by i-iivs nor 

Paul Shoret 39 

which we are required to apprehend them in thought or in the soul.^" It is often said 
that souls take the place of ideas in Plato's later period. This is a complete miscon- 
ception of Plato's thought and style. It is quite true that he could not confine the 
predicates of true or absolute Being to the ideas. God is, of course, true Being, and 
in religious and metaphysical passages need not always be distinguished from the ideas 
taken collectively. Both are invisible, eternal, intelligible. In the Timceus space also 
is reluctantly treated as a kind of eternal being. The Sophist tries to show that 
" being " is amenable to human logic and cognizable by finite minds. This involves a 
contradiction for all except consistent relativists who renounce pure Being altogether. 
This Plato could not do, for, not only in the Parmenides, but in the late Timceus, he 
retains absolute Being for metaphysics and religion. In the Sophist he shows that for 
human logic it is as impracticable as absolute not-Being, To be known and talked 
about it must come out of its isolation and enter into relations — act and be acted upon. 
Being is therefore temporarily defined against the extremists of all schools as the 
power and potentiality ^^ of action or passion, and the contradiction is smoothed over 
by the equivocal use of "true being" to denote both the metaphysical and the reli- 
gious noumenon — the ideas and God. True Being as God obviously possesses life, 
thought, motion, soul, and true Being as the ideas borrows so much life and motion as 
will explain their intercommunion in finite thought.^*' But the definition, its purpose 
served, is never repeated, and pure transcendental being reappears in the Timceus. 
That the ideas still take precedence of souls appears distinctly from Polit, 309 C, 
where it is said that fixed opinions in souls are a divine thing in a daemonic thing. 
The same follows from the creation of the soul in the Timceus, and the hierarchy of 
elements in the good (Phileb., 66) where pure ideas precede vovf. ^" Politicus, 269 D, pre- 
sumably implies the ideas ;"°' 285 E ff. unmistakably aflBrms them. What other possible 
interpretation can be put upon the statement on tois fiev t&v ovtoov paBioK KarafiaOelv 
aiadrirai Tivet o/jLoioTtjTe^ 'rre<j)VKacnv? These ovra are plainly ideas of material things, 
of which material things are likenesses. But ra TinimTara (justice, good, etc., Phcedr., 

^^^Sophisti 250 B, rptTov apct Ti TTctpa Tavra to ov ev Tfj i^xB ^*' e<TTos of Parmenides (or his followers at Megara or in 

TtOets, Qf, 243 C, ovx ^TTOv Kara to ov ravrov touto iradoq eiAjj- the school — ovSev yap TavTT[} Sta^ipet) as well as the TrdvTa pet 

^6t£s ev Tp ^xff. Cf. oiJ.o\oyritLaTa . . . . ey tjj ruLerepif tpvxfit of Heraclitus for which he felt less sympathy. Cf. ThecE- 

Thecetet., 155 A, from which Lutoslawski, p. 383, infers that tet, 180, 181, 183 E, 184 A. 

the ideas are subjective notions I 267 See Zellee, pp. 689, 690, who seems to deny the con- 

265 247 E, fivca/iis probably includes both. tradiction altogether, and pp. 696-8, where he argues that 

266 The entire passage betrays embarrassment. Toadapt the SopMst is early because Ufe and causality are never 
"Being" to the necessities of logic, Plato is obUged to fsa'n attributed to the ideas, and do not belong to them 
deny of it (248 DE) what in Tim., 38 AB, his feelings require '"^ Aristotle s representation. Space fails to enumerate all 
him to affirm. He treats ycyva.^Kea»a. as a Tra-rxe.. which P°'"*= °f tr^T I^ ^°f "?'^f'™°« from Apelt s subtle 
Zei,i.eb (p. 652), as a true Aristotelian, thinks a verbal fal- f'f'.f ^^fSophtst (B^trage). He points out that the 
1 T i.1- -1 ojft A V • ^ / 3f* ^- definition of ov is directed mainly against the materialists, 
lacy. In the crucial passage, 249 A, he uses auTo (|iti)6eji)i' , ,, ^^ .. ^ . .... . . , „ ■ 

■ -^ u- u J *.*.*.- ™ * 4.1 .'J and calls attention to to-tos ei? vCTTcpoi' cTepov ai* i^iai-eiTj, Heis 

auTo) which draws our attention away from the ideas. ..^. , . iti^., i, • t jj- 

... . ii 'v A J 1 J ■ J .. u -i. » 1 1 . right in denying that Plato s views changed, and in mini- 

And having attributed soul and mind to 'it, he merely ....:"„ ... ^ ^i -V ..■ « ,i 

..,,.. , . . . . , „ * u • mizing the significance of the apparent attribution of life 

infers that, since these involve /tii'iiiris, icnT)<r« must be in- ^ ^, ., %...,. ,. t , ,• -^ ^ .. 

, J , . / V v/-, u 11 J T„« „ „■.!„ „„ „ to the ideas. But he errs when he seeks an explicit state- 

eluded among ofTa (which CampbeU, aa toe., regards as a ,,.,. ,, ,., ,, ^,. 

, 1 » ,1 \ T>i • 1 v i ■ !•!.•„. *„..,„ ment of it in other dialogues and for this purpose presses 

formal fallacy). Plainly, whatever implications we force , j . m- cot. *•«-«- 

Til i , J . - . • i. i. i.^ 11 ..„ avTTVov Avcrtc, Tim., 52 B. 

upon Plato s words, his purpose here is not to attribute ^ ' ' 

soul to the ideas, but to remove from the path of logic the 268 to Kara ravTa koI uo-avTus ex^uv, etc. 

40 The Unity of Plato's Thought 

250 B, offa dWa Tifiia yfrvxal's), have no copies in the world of sense, and must be appre- 
hended by reason. This is precisely the doctrine of Phcedrus, 250 BCD and 263 A B, 
and ought to end controversy.'''^' We have already seen that the Philebus bids us 
assume ideas and disregard the difficulties of the Parmenides."'' There is no hint 
that they are only concepts."' We may assume, then, that the language of 58 A, 59 0, 
and 61 E implies the ideas. "^ 


Supposed variations in Plato's psychology have been used to determine the evolu- 
tion of his thought and the relative dates of the dialogues. The chief topics are: 
(1) the immortality of the soul; (2) the unity of the soul, or its subdivision into 
faculties; (3) the general argument that the psychology of the "later" dialogues is 
richer and more precise than that of the earlier. 

1. The immortality of the individual soul is for Plato a pious hope,"' and an ethical 
postulate,"* rather than a demonstrable certainty."" He essays various demonstrations, 
but nearly always in connection with a myth, and of all the proofs attempted but one is 
repeated. In the Apology Socrates, addressing his judges, affects to leave the question 
open."° But we cannot infer from this that the Apology antedates Plato's belief in 
immortality. For, to say nothing of Pythagorean sources of inspiration, he had pre- 
sumably read Pindar's second Olympian with approval; and Socrates's language in 
Criio, 54 B, is precisely in the tone of the Gorgias and the Phcedo.'^^^ The Meno™ 
assumes the immortality and the prior existence of the soul to account for a priori 
knowledge. The Phcedo presents a complicated proof or series of proofs. The Sympo- 
sium seems to recognize only the subjective immortality of fame, and the racial immor- 
tality of offspring."' The "early" Phcedrus and the late Laws alone agree in a proof 
based on the conception of the soul as the self -moving."*" It is easy to foresee the 
hypotheses which an ingenious philology will construct from these facts. Krohn, Pflei- 
derer, and Eohde gravely argue that Book I of the Republic must be very early because 
the aged Cephalus neglects the opportunity to supplement his citation from Pindar with 
a scientific proof of immortality. Horn tells us that the Phcedrus represents the first 

269Fora;'a)i>'iI(7ismthePoiiKcMsc/. m/ro, p. 44. Phmdo, 115 DE; and with the idea, 959 B, that the only 

2™ See A. J. P., Vol. IX, p. 279. Po^Seia at the bar of Hades is a just life in this world, cf. 

271 Ldtoslawski, p. 467, mistranslates, or, if he prefers, Gorg., 522 C D, 526 E ; Onto, 54 B. 

misinterprets, 15 D : " the nature of thought requires the 275 Phcedo, 85 C, to iikv o-ai^e; tlSiuai 4v tu vvv fiia ^ iSvua. 

union of notions into higher units, and this constitutes an toi' elmt ^ irayx'''^'™'' ". Cf. 107 A B ; Tim., 72 D ; Memo, 

eternal necessity of the human mind. " Cf. supra, p. 36. 86 A B ; Phcedr., 265 C. 

272 TTji' yap TT€pl TO or kol to orTW? icai to Kara ravTov aei 276 4Q Q, Cf. also Pkcedo 91 B. 

, , '. , , , . . . ^^^Cratylus, 403 DE, implies the doctrine of Phcedo, 

ra a-VTcL ilcravTcos afliKToTnTa exovTa.— i) «e eiri Ta j«)Te yi,yvoit.eva. (57 oa 

w>7Te aTToAAu'iLLera, «aTo ravTa 6e «ai wcraUToj? ovTa aei. Cf. 62 A, ' ' 

{t7fl S"! r^ 
aiiTrjs irepi 6c«aio<n<V7js o Tt eCTTi. 66 A, rT)V aiSiof .... ^vtriv, "-1 ^■ 

For the ideas in relation to the method kot' elSii Tejireiv, 279 207 D, 208 B. Too much is made of this, for the same 

and a fuller discussion of the m ov fallacy, see infra, inference could be drawn from Laws, 721 and 773 B. The 

Part II. popular belief in Hades is implied, 192 E, and there is even 

273 Phoedo, 114 D, XP^J '""^ ToiauTa iinrirep iirqSeLv eavT^, a hint, 212 A, that the philosopher may be immortal: elirep 

274 Sep., 608Cff.; Laws, 881A, 967DE, 959AB; with tc? a^Aij. i^Spcuirui- iewaTij. ical KeiV^. 
Tor Si OVTO. iiiJ.aii' eftaffTOr oPTujt iBavdrov [eirai] iliyxv", cf. 280 Pftajdj.,, 245 C ! iaws, 894, 895, 

Paul Shobey 41 

youthful enthusiastic apprehension of immortality, the Symposium expresses the mood 
of sober manhood content with this life, while in the Phcedo old age. waiting for death, 
craves a real immortality. According to Thompson, the Meno reserves the proof of 
what it merely asserts; the Phcedrus outlines a general proof, the Republic later 
attempts another ; the Symposium, dissatisfied with all so far achieved, ignores the sub- 
ject ; and finally the problem is taken up seriously in the Phcedo. Zeller, on the other 
hand, while holding that all the proofs are substantially identical, thinks, as we have 
seen, that the Republic refers to the Phcedo, and is also later than the Phcedrus. But 
to Lutoslawski it is evident that the proof given in the Phcedrus and repeated in the 
Laws is the latest. And he also can discern that the Symposium, in the first flush of 
idealism, could dispense with the personal immortality of the Gorgias, but that later, 
when the theory of ideas had grown familiar, Plato undertook in the Phcedo to affiliate 
upon it the old doctrine of immortality. 

Hardly more profitable than these arbitrary speculations is the analysis of the 
separate arguments. Broadly speaking, Zeller is right in saying that they all amount 
to this, that it is the nature or essence of the soul to live. But this general truth 
becomes a fallacy when employed to identify absolutely the distinct arguments of the 
Phcedo, the Republic, and the Phcedrus. The gist of the argument in the tenth book 
of the Republic is a fallacy employed also in the first book (353 D E), the equivocal 
use of the aperi] or specific excellence of the soul in relation to its epyov, its function 
and essence. In both cases the epyov is defined in terms of mere life-vitality, while 
the aperij is referred to the moral life. But in so far as the epyov or essence of the 
soul is mere life, its aperi] is intensity and persistency of life — not justice.^" Simi- 
larly the Phcedrus and Laws, identifying life with self-movement, prove the eternity 
of the principle of motion, and assume it to include moral and intellectual qualities.^'^ 
But there is a certain pedantry in thus scrutinizing these arguments. Plato's belief 
in immortality was a conviction of the psychological and moral impossibility of sheer 
materialism,'*' and a broad faith in the unseen, the spiritual, the ideal. The logical 
obstacles to a positive demonstration of personal immortality were as obvious to him 
as they are to his critics. If we must analyze the arguments of the Phcedo, the 
analysis of Bonitz is, on the whole, the most plausible.^" They prove, at the most, 

281 Of. the equivocal use of ipiiovia in Phcedo, 93, 94, to objections by establishing the inherent immortality of the 

denote the composition of physical elements that, on the soul as a form that always involves the idea of life. I may 

hypothesis under examination, is life, and the harmony of add that the fallacy in this ingenious argument may be 

spiritual qualities that is virtue. analyzed in various ways. In 103 B it is said that aurb to 

282 Laws, 896 C D. 

evavTiof, as distinguished from ra txovra to, ivavTia could 
never admit its opposite. Auto to evavriov is then sub- 

283 Laws, 891 C, KivSmevei. ya/i 6 \iyai/ TaSTa iriip Kai S8io(> divided into TO iv ij/xii' and TO ei- TJ) ^vaei. This seems to 
Kill Y^y «ai ie'pa TrpuTu ^yeiaflai Tilv irai-Tior eti-tti. Cf. Phileb., yield three things: the ideapcrse, the idea in the particular, 
30A; TheoBtet., 155 E, 184 D; Sophist, 246Aj Tim., 51 C, and the particular as affected by the idea. (C/.SMpra,n. 252.) 
r, TauTa, oirep «ai ^AeTrojiei- .... i^ova iml ToidvTtiv IxovTa bqj there are really only two things: the idea, and the 
aA^Seiav. particular affected by the " presence " of or " participa- 

28i J. €., the argument ex rStv ivavrCtav Ta ivavria, 70 E ff., tion " in the idea. How the idea can be at once in itself 
proves merely that the state of the soul after death is the and in the particular may be, as we have seen, a mystery, 
same as that before birth. The argument from irajinjo-w. But it does not justify the duplication of the idea, which is 
73 ff., supplements this by the proof that before birth the a device employed here only, and presumably with full 
soul possessed intelligence. The final argument meets all consciousness, for the purpose of the argument. For by its 

42 The Unity of Plato's Thought 

the immortality of soul, not of the individual. This Plato presumably knew, but 
we cannot expect him to say so by the death-bed of Socrates or in the ethical myths, 
which obviously assume individual immortality.'''* But neither this unavoidable funda- 
mental ambiguity nor the fanciful variations of the eschatological myths convict Plato 
of serious inconsistency, or supply any evidence for the dating of the dialogues. 

2. In the Republic Plato bases the definitions of the virtues and the three classes 
of the population on a tripartite division of the soul, which he warns us is not demon- 
strated absolutely, but sufficiently for the purpose in hand.^'° A poetical passage of 
the tenth book hints that in its true nature the soul is one and simple, but that we 
cannot perceive this so long as, like the sea-god Glaucus, it is disguised by the accre- 
tions of its earthly life.^" The tripartite division is embodied in the myth of the 
PhcBdrus, which, if we pedantically press the poetical imagery ,'*' implies the pre- 
existence even of the appetites.^'' In the Timceus the immortal soul is created by the 
Demiurgus, the mortal, which falls into two parts, spirit and appetite, by his minis- 
ters.^" Here the tripartite division is subordinated to a bipartite, as Aristotle would 
have it.^"' But we are explicitly warned that the revelation of a god would be required 
to affirm the absolute scientific truth of this division, and to distinguish precisely the 
mortal from the immortal part.^''* In the Laws the question whether the dv/j.o'; is an 
affection or a distinct part of the soul is left open.'"" As Aristotle says, it makes no 
difference for ethical and political theory.^* The Phcedo, attempting to prove immor- 
tality, naturally dwells rather upon the unity of the soul, as does the tenth book of the 
Republic. But it distinguishes, quite in the manner of the Republic, the three types 
of character, the <j)i\da^o<f)OV or (jyiXofiad'^, the <pi\apxo<; or ^iXdrtyito?, and the (J3i\oa-co- 
fj,aTO<; or <f)iXoxfyi]fJ'aTo<;.'^^ Phcedo, 79 B E, does not affirm that the soul is absolutely 
simple and uncompounded, but that the body is more akin to the composite, and the 
soul to the simple and unchanging. The contradictions found by Krohn and Pfleiderer 
in the psychology of the Republic, or between the Republic and Phcedo, on this point, 
are sufficiently explained by Hirmer.™" From all this it appears (1) that Plato 
affirmed nothing dogmatically with regard to the ultimate psychological problem. 
(2) That his primary classification was the distinction between the pure reason and the 
lower faculties subordinate to reason and dependent on the body. (3) That for ethical 
and political theory he found most helpful the tripartite classification — reason, spirit, 

aid the life in the individual is posited as an intermediate Plat, Forsch,, p. 33, says that Rep., X, must be later than 

entity between life per se and the living individual, and Phcedrue, for in the Phoedms immortality belongs to all 

pronounced immortal because, like life per se, it will not three parts of the soul 1 

admit its opposite. Another way of putting it is to say that, 290 34 B c 69 C ff . 

in 106 Eff., aWvaToi/ is equivocally used for (1) that which 00, tjui >t' -. .0 n • . , . . - , ... 

does not admit death (while Ufe is present), (2) that which ^. l^.'^^' ^'"'^ ^' ^^' ^' °'°'' " '""' "''"l""' '"'^'" '^'""' " '' 

does not admit death at all. ^ * 

38! Gorff., 524 ff.;iJep., 614 ff. Of. Laws, mBC; Tim., 292 72Di c/. PA<Bdr., 248 A. 

41 D, yf/vj^ai taapid/iovi ToU tttrrpots, etc. ^^^ ^3 ^1 «'" Vt irddo? elre n fitpoi Stv o 6vii.6s. 

286435CDff. 287 611 C - 612 A. 288246A1I. ^^* Eth. Nic., 1,13,10, oiSiv Sia(i><pei,Trfi6i to jrapai/. 

289 Natoep, Hermes, Vol. XXXV, p. 430, objects that the ^^^ ^8 C, 82 C. 

souls of the gods are tripartite and that the horses, though 296 " Entstehung und Komposition der Plat. Politela," 

in the procession, do not see the ideas I Scsbmihl, Nev£ JahrbUcher fUr Phil., Suppl., N. F,. Vol. XXIII, pp. 642, 643. 

Paul Shoeey 43 

appetite — which he also embodied in the myths of the Phozdrus and the Timceus. 
(4) That, while this classification may be profitably compared with the modern intelli- 
gence, feeling, will, it is beside the mark to criticise it as if it were meant to be 
psychologically exact and exhaustive.^" We cannot establish any fixed relation 
between the tripartite soul and the hierarchy of the cognitive faculties — wO? (vorjaK, 
eiriaTrfiirj), Bidvoia, So^a, 7ria-Ti<s, elKaaia, etc.'''' Plato sometimes treats the inerrant 
reason as a distinct part of the soul from the fallible faculties of sense and opinion. ^^ 
He sometimes associates sense-perception with sensuous appetite in common antithesis 
to the reason.™' But he also, when it suits his purpose, virtually identifies (true) 
opinion with reason, in opposition to the impulses of instinct and appetite.'"' The 
dviw<i, though associated with opinion,*'^ cannot be assigned with it to a distinct part 
of the soul.'"' Nor can it be identified with the "feeling" of the modern psychologist. 
The will as a faculty distinct from the impulses of appetite and the judgments of the 
reason has no place in Plato's system. (5) That we cannot fix the time at which the 
notion of the tripartite soul first occurred to Plato, nor may we use apparent variations 
in the mythological dress of the doctrine in order to date the Phcedo and Phcedrus 
relatively to each other or to the Republic. 

3. The chief changes alleged in Plato's "later" psychology are: (a) the abandon- 
ment of avd/jLvrja-K; (b) a different conception of the relation of mind and body, more 
particularly as concerns the nature and seat of pleasure and pain; (c) a fuller and 
more precise terminology of the cognitive faculties and the degrees of knowledge. 
This later psychology must be sought chiefly in the Philebus. It is not enough to 
point out that the Philebus is especially rich in psychological detail. The subject 
called for it, and we cannot expect all the dialogues to be equally full in every topic. 
What is required is contradictions of earlier dialogues, or new thoughts not hinted at 
in them. And these are not to be found. 

a) The explanation of the ordinary psychological meaning of avafivrfai'; in Philebus, 
34 B, no more proves the abandonment of the peculiar Platonic doctrine than does the 
occurrence of the word in that sense in the Republic, 604 D. The Phcedo itself treats 
the avdiJLVT]ai<i of the ideas as a special case of recollection and association of ideas gen- 
erally, and employs the consecrated phrase tovto S' iarlv avdfivqai'; of an example that 
fits the definition of the Philebus.'^ Plainly all recollection of the ideas is avdjivrjai'^, 
but all avdfiVTjaK need not be recollection of the ideas. Moreover, as the word occurs 
without the doctrine in the Philebus, so we find the doctrine without the word in the 
Politicus. As the point has been overlooked, it is worth while to dwell upon it. Every 

297 See JowETT, Vol. I, p. 410 ; Zbllee, p. 846 ; Lnioa- 3M Phileb., 60 D j P/wjedi-., 237 D ; infray p. 4S, u. 357. 
liAWSEI, p. 278. 302 This is probably the meaning of a\r)9tvT)is Jof tjs eraipot, 

298 The imagery and terminology of Sep., 511 D, 534 A, Phcedr., 253 D, despite the antithesis aXa^oreiaj eraipos. 
belong to the literary machinery of the Bepublic, and are oXijSivii is used of So^a = opinion in Themtet., 187 C ; Phileb., 
not to be pressed. 37 B. 

299Bep., 478AB, 602E-60SA, to iropi ri (nVpa apa {of i^ov 303 In Km., 37 B C, «of ■" and irio-Teit belong to the circle 

Tqs i/fux^s Ty Kara to ^e'Tpa ouk av etij toutoi'. of the Barepov in the immortal soul. 

800PAcedo,65, 66. 304 73D. 

4:4 The Unity of Plato's Thought 

man, we are told, knows all things as in a dream, thougli he fails of waking knowl- 
edge.'"^ This at once recalls the fjt^fuiOtjKviai; tjj? '^v)(rj<: airavra of the locus classicus 
on avdfivj)m<;, Meno, 81 D. In the Meno, too, it is said that this knowledge is at first 
dreamlike, bnt is converted by the elenchns into true science.*" The Politicus goes on 
to show, by the use of Plato's favorite illustration of letters or "elements,'""' how it is 
that, despite this antecedent knowledge, we go astray, and how in the study of complex 
and difficult things the right use of example and comparison will enable us to recognize 
the identity of the same form or idea everywhere, so that we shall have a waking and 
not a dreamlike knowledge. *"' Children, knowing their letters in some sort, distinguish 
them rightly in easy combinations, but blunder in long hard syllables, until by compari- 
son with the easy they learn to recognize the same letter everywhere. So our soul, 
similarly affected by nature toward the elements of all things (the ideas), sometimes and 
in some things is settled and fixed by truth concerning each one, but at other times and 
in other things is driven to and fro among them all, and of some it somehow forms 
right opinions among the combinations, but fails to apprehend these same things when 
transferred to the long and difficult syllables of facts. Not only the general drift, but 
the language and imagery of this passage must be understood of the recollections of 
the ideas. The phrase Tainov tovto fifiwv r] "^vxt) <f>vcr€i wepX to, toiv ttcivtodv crrot^^eta 
ireiTovdvla does not refer mainly or solely to our liability to error, as might be sup- 
posed from Campbell's "is naturally liable to the same infirmity," or from Jowett's 
"has the same uncertainty." It refers to the whole preceding comparison of which the 
starting-point is that the soul knows all things in a sense, even as the children know 
all their letters imperfectly. That this is the meaning of (jivaeu .... ireirovOvla appears 
further by comparison with Phcedrus, 249 E, "jraa-a fiev avOpayirov y^vxri (j>va-ei, redearat 
ra ovra. The doctrine of avdfivrjai<;, then, repeated in the Politicus, is not abandoned 
in the Philebus. This conclusion might have been affirmed a priori. For " recollec- 
tion," once indissolubly associated with the ideas and the pre-existence of the soul, 
would not be given up while they were retained. But pre-existence is assumed in the 
Laws,^' and the ideas, as we have seen, occur in the Politicus^^" and are reaffirmed in 
the TimcBus, which also implies the soul's prior knowledge of all things, in language 
recalling the Phcedrus and Politicus.'^^ 

h) The general problem of the relation of mind and body is involved in that of 
immortality and the parts of the soul. As we have seen, the Timceus, though it assigns 
separate seats to the mortal and immortal soul, declines to dogmatize without the assur- 

305 277 D, Kivfivfeuet yap 7}fiwv CKOtrTos olov ovap ct£fa>9 airafTa 306 Meno^ 85 C, Sarirep ovap apri KeKLVTiVToi. al Bo^ai a^rat. 

a.i ira\i.v i>(T7rep vTrap ayvoeZv. RITCHIE, p. 143, misapprehends 307iJep«6., 402AB; cf Soph., 253A; Phileb., ISC; Thece- 

this passage when he associates it with the " lie of approxi- f^f 201 B* Tim. 48 B etc. 
mation.'^ We must nse examples, not because in difficult 
matters it is permissible to fall back upon "picture- 
thinking and symbolism," but because only by beginning 
with easy examples can we learn how to convert our dream- ^"*i ^*®' 

like knowledge into real knowledge. The yap introduces siogupra, p. 39. 

the whole parallel, of which the dreamlike knowledge of 311 41 E, t>|v toO iravTb? (Juffii- eStife 

all things is only the first point. 

308 278 E, rexvjl yv<iipt^ii,v, 'iva, iilrap avT^ ovct'paroc 

Paul Shoeey 45 

ance of a god, and the Laws leaves it an open question whether the parts of the soul 
are real parts or functions.'" Of the dependence of our cognitive faculties on bodily 
organs Plato knew as much or as little as we know.'" In the images of the wax tablet 
and aviary he anticipates all psychologies that explain memory, association, and recol- 
lection, and the distinction between latent and actual knowledge, by material analogies.'" 
But sheer materialism and sensationalism he rejects, for many other reasons"^ and 
because it fails to account for the synthetic unity of thought."* The senses are the organs 
through which, not the faculties by which, we know.'" Sometimes and for some purposes 
he exalts pure thought freed from all contaminations of sense."' In other moods, he 
recognizes that human thought takes its start from aiaOrjaK or immediate perception."' 
He points out that the contradictions of sense give the first awakening stimulus to the 
generalizing activities of mind.'^° He admits that our minds are too weak to attain to 
knowledge without experience,'^' and require the aid of concrete examples in order to 
apprehend difficult abstractions.'^ We can recover the prenatal vision of the ideas only 
by association with their sensuous "copies," or by strenuous logical discipline.'^' And, 
though knowledge is not sense-perception, sense-perception is the best evidence that 
we have of some things.'^* Only a very literal-minded criticism will treat these con- 
cessions as a contradiction of the apotheosis of pure thought in the Phcedo. 

Slightly more plausible is the claim that Plato contradicts himself in regard to the 
nature and seat of desire, pleasure, and pain.'^^ The "early" Qorgias and the "late" 
Fhilebus explicitly affirm that the soul, not the body, is the seat of desire.'^' The Philebus 
adds the psychological reason that desire is dependent on memory.'" The Philebus 
further explains pleasure and pain as mental states arising from changes in the body 
sudden enough or violent enough to affect the mind and pass the threshold of con- 
sciousness, in modem phrase.'^ Pain results from movements unfavorable to the 
"natural" condition of the body, pleasure from those that preserve or restore the natural 

snSupra, a. 293 ; cf. also iJep., 612A, elre iro^veiSiit elre 322poZtt., 277D, Cf. Phcedr., 262 C, i^iAils jrus Aeyofiev 

ftoyoeiS-^^f Phcedr,, 271 A, oiiK exovre^ iKava irapa^elyfiara. 

3i3Pft(Bdo, 96BC, TT&Tepoi' TO al/ii kariv a 4ipovovii.iv, ^ 6 323 pftcetJo, 75A; PoZtt., 286A; Bep., 533A, icai on i] to5 

artft ri To Trvp, etc. Note the irony of the whole passage. diaXe'veo-flm ivranw fidi/j) av ^•ivii.^v iiiireipif ovTi !iv viv ii) 

31* Die<Bte<., 191 D fl. (cf. FTwEdr., 275 A, TvVio^), 197 D, ,' . . ',. '- . ,,,,.- .," 


^^^TheCBtet., 201 B, Stv ifiorTt Ikovov itir^v elBevat a\Xtit<; Se 

n^Phcedo, 80B, 96; Phileb., 30; Tim., 51 C; Laws, 889. ^^^ ^^^^^^ 234D, koX «.i raen>,v ivay.aiop.ivov, ivapyC, 

816 Thecetet., 184 D. itttdirTeaBat, Ttiiv ovriav. The whole passage is in seeming con- 

„ „ „ „ tradictionwiththethonght of PAoedo.lOOA, andJJep.,473A, 

3" TAecetet., 184 C ; Phcedo, 65 D, 79 C ; Tim., 67 B. ^^^^^ ^^^^^ (thought) come nearer to truth than deeds. See 

ns Phcedo, 65 C (cf. Theostet., 187 A), 66 A, eikiKpivel rg also Jfeno, 87B. 
SMvoCa; 67 C, t6 x^P'^e''' o " liiKurra iiro roB aii/iaTot rhv 825(Jrote, Jowett, Mr. Henry Jackson, and others. 

'h'xvv. HOBN, who rejects the Philebus, says (p. 380) that it assigns 

319 Thecetet., 179 C, to iraphv «a<rTo) 7ra9o! e| &v ai aiffS^crew desire to the soul, but pain and pleasure to the body. 

Koi at KOTO TOVTO! Sofcu. Charm,, 159 A, oi(r9i)<nV nva iropex""! 326 Gorg., 493 A, t^i 6e "/""X^s toOto iv <J imivpiiai. eiffi. So 

ef fls fidfo av Ti's aot nepl auT^s €11?. Phileb., 249 B, ex iroWav Tim., 69 C. 
I'ov aXv6ri(rvav eis €^ AoytCTjitoi ^vvaipovti.€VOV. 

320 Bep., 524 B C ; Thecetet., 186 A B. 3jg 33^ ^^ ^3 g q p^ jj^p _ ^j C, 584 C, ol ye Jii toO aap-a- 
3'a TheCBtet., 149 C, oti i) avtpiairivri glials aaBevtaripa ^ TO! eirl Tiji' iivxnv Ttivovuai. «oi \ey6p.tvai. ^Sovol. Cf, Laws, 

Ka^tlv rixvv" "^ «■' if «te'P»5. 6'3 A, jiexP' T^t •I'vxv^ ; Tim., 45 D (of sensations). 

46 The Unity of Plato's Thought 

state.'^' This is also the doctrine of the Timceus, and it is not contradicted anywhere. 
In ethical and religious discussion, however, it is natural to identify the "soul" with the 
higher intelligence, wO? or immortal soul, and to speak of the pleasures of the mortal 
soul which come through the body and are necessitated by the body as pleasures of 
the body. And Plato, though usually scrupulously precise,^** occasionally permits 
himself this inexact way of speaking. The Philebus enumerates three kinds of mixed 
pleasures and pains: (1) merely mental, as in the pleasurable-painful emotions; (2) 
merely bodily ; (3) those that arise when pleasure of mind accompanies pain of body, 
or the reverse.^' In a few cases the "bodily" pleasures are spoken of as if they were 
literally in or of the body.™ But Plato was justified in assuming that only a careless 
or captious reader would misunderstand him. For hardly three pages back he had 
explained that bodily states produce pleasure and pain only when they cross the 
threshold of consciousness. ^'' There are also two or three cases in the Phmdo. In 
the first the phrase "appetites of the body" is used in a highly wrought, ethical pas- 
sage precisely as it might be employed by a modern preacher, with no implication of 
psychological doctrine.^'* The second occurs in the refutation of the hypothesis that 
the soul may be a "harmony" of material states or elements. To refute this objection 
Socrates employs the very argument used in the Republic to distinguish wO? from 
iiriOvfiia and Ovfwi;.^ The soul cannot be identical with that which it rebukes and 
controls as a superior. The soul, instead of being controlled, vtto t&v tov aafiaTOi 
vad&v, is master of them. Therefore it cannot be a "harmony" composed of them. 
The appetites are treated as material iraOrjiiaTa in order to refute, in its own termi- 
nology, the hypothesis that the soul is a composition of material iraOrifiaTa. The 
argument would lose its force if stated in the terminology of the Republic. If the 
tripartite soul were explicitly recognized, it would be necessary, first, to decide which 
parts are to be immortal; secondly, to prove directly, and not by the equivocal substi- 
tution of "bodily" appetites for states of matter, that the wvs or soul cannot be a 
harmony of material elements. For these reasons, in the Phcedo, soul, tacitly identified 
with vov'i, is opposed to body as a whole, including the appetites. But the literary 
and aesthetic necessity of this way of speaking having once been perceived, we cannot 
treat it as a contradiction of the psychological truth clearly stated in the " earlier" 

s^Phileb., 31D £E., 42D; Tim., 64 CD, 66C, 68A. Im- £» to! koh-u .... -yeVei is merely preparatory to the explana- 
plied perhaps " already " in Cratyl,, 419 C, ij re Auwt) airb t^s tion that they are the psychic correlates of beneficial or 
6taXv(7€coy Tou (TwiuaTos. Aristotle, Eth. Nic, 10, 3, 6, contro- harmful changes in the body. It is obviously no contra- 
verting the doctrine that pleasure is a -yeyco-ts, says : et diction of the reference of ^Sov^ to the a-neipov in 31 B. Qf. 

5^ effTt TOU Kara, tftvtnv avan-X^ptocris rjSovri, ev i^ ij avatrkiqfKocriit A, J P., Vol. IX, No. 3, p. 284. 

toOt' if Kol riSouro- to ai,ii.a ipa- ov «o«e: Se, where oi SoKtZ 333 43 g q C/. 33 D, 9« Ti;- irepl to a!i, .... ira9i,/iaT<oy 

expresses as often Plato's opinion. ^i ^l„ ^^ ^^ aci^aTi /carao-^e^vvjie^a irpl" «t' tSj^ ^vxv^ Jief eXOeli.. 

3^ophileb., 39 D, rS,v 6ii tov o-ol^aTO! r,SovS,v. So 45B, This is the doctrine of Tim., 64 ABC, and it is " already " 

Phcedo, 65A; Tim., 64A; Sep., 584C, 4S5D; Phileb., 45 A, implied in TAeoetef., 186C, oo-a 8ia toS o-ojuoito! irae^finTu eirl 

oi irtpl TO <T(una. So Pftcedr., 258 E. Cf.,Cratyl.,404:A.; Bep., rriv xjnixriv reivei.. PAilei>., 55 B, explicitly affirms that pleas- 

442 A; Tim,, 64 A; Phileb., 41 C, to trUiia ^v to jrapexo/iei'oi'; ure is in the soul only: wit ovk aAoyo;' inTi. /j.riSei' iyaOhv elfoi. 

Rep,, 584 A, to ye ^5v ev ^XV yi-yyofievov; 442 A. .... it\i}v er i/'i'xn "at evTav9a ijSovtjv ij.6vov, 

331 47 E-50D, 46C, 47 C D. 3^^Q&C,KaX yap iroKenov^ Koi araaets KaX ^a;^a; ovSel/ oAAo 

332 46BC, 50 D. So Prodicus in ProtOff., 337 C. The irapexei i to o-ijia ical oi tovtou eiriSufiial. 

statement, Phileb,, 31 B, that pleasure and pain originate ^^^Phcedo, 94B ft.; Pep., 441 B, 390 D. 

Paul Shorey 47 

Gorgias and "later" Philebus. One might as well argue that the tenth book of the 
Republic antedates or abandons the tripartite soul because the doctrine is ignored in 
the proof of immortality attempted there. 

c) Lastly it is sometimes affirmed that the later dialogues show an increased preci- 
sion in the use of psychological terminology. In fact, however, Plato's psychological 
vocabulary is nowhere technical. He is content to make his meaning plain by the 
context. Nor can we find in Spinoza or Kant or in any modern text-book the consist- 
ent precision that is sometimes demanded of Plato. There is no modern terminology 
which sharply discriminates mental states that are or are not supposed to involve the 
element of judgment and belief. There is none that shows independently of the 
context the precise line intended to be drawn between sensation and perception, 
or distinguishes revived and compounded "images" from "images" regarded as 
immediate impressions. We cannot, then, expect Plato to emphasize distinctions not 
needed for his immediate purpose, but if we bear this in mind, we shall find no serious 
inconsistencies or significant variations in his use of such terms as a'iaOqaL'i ho^a and 

M(Tdr](Tt<i is any immediate sensation or perception or consciousness including 
pleasure and pain and Locke's inner sense.^^ As sense-perception it is rightly said to 
involve judgment,'^' and so issues in ho^a, opinion or belief.''* The word So'^a may 
be used in this neutral, psychological sense ; it may be taken unfavorably to denote 
mere opinion as opposed to knowledge, or favorably when true opinions and beliefs are 
set in antithesis to the appetites and instincts.'*' These shades of meaning arise 
naturally out of Greek usage, and would call for no comment if they had not been cited 
to convict Plato of inconsistency or change. The mental process that terminates in 
the affirmation or negation that constitutes ho^a may be expressed in words, Xo'70?,'*' 
or take place in silent thought. In the second case it is Sidvoia — a discourse in the 
soul.'*' Aidvoia, then, mere or silent thought, may be opposed to speech'*^ or to 
thought accompanied or interrupted by sensation.'*' It is thus often a synonym of 
pure thought.'" But the Republic, in default of a better term,'*^ employs it to denote 

336rAecctet.,156B,186DE,152BC; Philel>.,UK; Charm., SMPhileb., 38E, koX Xoyos Si, yiyovtv oCtu! 'o totc Sdlai- 

159 A. eicoAoCfie^. 

,„„„ ,, , 3il Phileb., 38 D; Thecetet, 189 E, 190 A. Soph., 263 E, 

33' Rep., 523 B, (is iitai-cos viro ttjs aiaOriireai Kpivonera. ■' '"">"•' "" . _ . ' , , 

Phileb., 38 C, 7roAAa«« Mpr. .... ^ovl^^eac Kpiv..v «<u,. S^ S--""- ''^■' «" '^"VOI Ta.rd. ,rX,. o ^v, t„ ^v^r,, npo, a.r,. 

TaS8' anep opf. This is not quite the modem psychologist's SiaAoyo!, etc. 

recognition of the judgment involved in perception, but it mSoph., 238 B, 261 A. 

leads up to Aristotle's characterization of sensation as 3*3 TAecefet, 195 C D ; iSep., 511 C, iiavoi^i niei' .... oAXi nrj 

Svlla^il.v <rviii<tivTov KfiTiKiiv. Analyt. Post., in fine. al(i9r,(re(Ti.v. In Phoedo, 73 D, it is the (memory) imagination 

asaphiUb., 38 B, « ij-v^jm' " «al aUrtijireio, Wfa. Phcedo, of modern psychology: «al er rf SmvoU l\aPov to eKos toO 

» TouTOv (sc. the senses) Si y^t-oito m-^Vi ««■' Sof». Charm., '""««.• in Bep., 603 C, it is the mind, including higher and 

159 A, al(re,,i7tv . ... el it S6^a. In Thecetet, nOB, iA.,95 lower faculties. 

SiavotaF .... ^evSri So^av, Sidvoia and fiofa are virtually 3*4 Phaedo, 66 A, elKucptveZ rjj Star oi'o ; 65 E, avro etLaxnov 

synonyms. Stavoriijivtu,. In Thecetet., 195 DE, we pass from an image 

339pft;7e6.,60D, n^wiv «al ^poFijiii- «al a\-r,^ {dfar T^s of a man, Sr «ia..oovMe9» (^di-o^, ipine.- S'oC, to abstractions as 

avTt,l ISia, rMiLtvK. Phoedr., 237 D, .V.(.vt« iniOv^xia .... ri ev««a S ^tiSJi- aXAo^ Siai/oelrai tw ; cf- Sep., 526 A, !>v Jiwoi)- 

imKTiros S6(a. Tim., 11 B. In Thecetet., l&l A, So(aieiv is fl^^at ^dvoi- iyx"?"- 

almost the pure thought of Phcedo, 65 C. 3»5 533 D, ou irepi m6p.aTot i)i^i<r(3^Ti)(ri5. 

48 The Unity of Plato's Thought 

the processes of mathematics and the sciences, which are inferior to the pure thought, 
vow, of dialectic, in that they depend on sensuous imagery and hypotheses.^*" 

Plato describes memory images,''" and images of "imagination."^ But he has 
no term for imagination as a faculty intermediate between abstract or verbal thought, 
on the one hand, and sense-perception, on the other. For <f)avTaa{a takes its color 
from (^aiverai and (pavrd^eTai, which include all forms of opinion and illusion, and it is 
often merely a disparaging synonym of So^a.^^ But ^aiverat, though applicable to 
any notion that appears true, is most naturally used of the appearances of sense, and 
so ^avracTia is preferably the form of ^o^a that accompanies sense-perception,^™ and 
may be defined as a-vfifu^i^ ala-Orjaeco'; Koi Sd^a?.^^' Pure infallible knowledge as an 
ideal must be sharply distinguished even from true opinion.'^^ Strictly speaking, it 
cannot be defined,''^' and is unattainable in this life.^* Poetically it may be described 
as the vision of the ideas, and we may be said to approximate to it in proportion as we 
"recollect" the ideas by severe dialectic.^' Practically knowledge is true opinion, 
sifted and tested by dialectic, and fixed by causal reasoning.*^" "True opinion" may 
be disparaged in contrast with the ideal, or praised as a necessary stage toward its 
attainment.**' It is a very mechanical criticism that finds contradiction or inconsist- 
ency here. 

There is no limit to the contradictions or developments that a false subtlety can 
discover in Plato's psychology. Most of them are by implication explained away in 
the foregoing summary. I will close with two or three further examples which must 
stand for all. 

Susemihl ™ argues that the Theceteius marks an advance on the psychology of the 
PhoBdrus because it includes Wahrnehmungsurtheile in SoKeiv or Bo^a.^^ But the 
Thecetetus itself elsewhere attributes them to aia6i](n<;, for only so could it identify 
Protagoras's theory with the definition aca-drja-c; = ima-Trifir]. As we have seen, the 
distinction is futile, for al<Tdr)<n^ may at any time be the modern sense-perception, 

S46 iJep., 511 D, 53i A. See Idea o/ Good, pp. 230 ff. i^ ThecBtet., 152C, i^avratria apa koX ai<r»i)o-ii niiTor iv tj 

snphileb., 39C; Ptodo, 73D; Thecetet., 191D, «cos Sr «'l>l^oZ, Kal ni^i toI, tolovto,,. Soph.,26iA,iTav m Kae' airnv 

iv^ TO elSoiXov avToO, eto. "'^^ *'' '"''*^''«"' ""^ri rm TO ToiovTo.- ail 7ri»o! ; i. e. it is here 

not a memory image, but a percept accompanied by belief. 
3^^ Phileb.^ 39 C, irept .... rSiv fteWovTutv; 40 A B, and 

the fantastic account of the functions of the Uyer, Tim., , , ^' ^"P^- ^^- ^enoe here 263 D, 4,avTa,rU, and Phileb., 

71 A B. Grote, expecting the modern atomistic order : sen- ^ ^' *'""-»<"*«™ (= imagmations or imaged expectations) 

sation, image, idea, judgment, is surprised that in PhiUb., ^"^^ ^^^^ *° ^'i™i' *'^"*^ ^""i falsehood. Modern atomistic 

39, memory and sensation first write \6yo, in the soul, and Psy^olosy sometimes conceiTes " images" as mere pictures 

that, secondly, a painter supervenes who paints images of i-^™!""? no affirmation or belief. Aristotle seems to ex- 

these Wyo. and the corresponding 6dfac. But it is charac- P^«^^ ^^^^ ™w ™ ^« Anima, 432a, 10, eVri «• i, ^.a^ra^U 

teristic of Plato to put the image after the idea, the word, ^"1""' *""■"' "»' «'""(>«<'""5- Bit in 428o, 12, thinking of 

and the judgment everywhere. Moreover, the images here P^"ebm, 40 A B, he says, ai «€ ■j.avraaia.K yCvovrai ai n\Uavt 

are not the primary images of perception, which are in- '''"' "'■ 

eluded in Plato's aIo-9rj<j-is, but imaginative visualizations ^^^ Tim,, 51 D E. 353 Thecetetus, infra ; supra, p. 43. 

of beliefs and hopes. In the mature human mind this is 36* Phc^do, 66, 67 ; Laws, 897 D, w vovy ^oTt e./,,Tor5 5^^a<r.K 

probably the real order: (1) sensation (perception), (2) hijioiitvai., 
faint verbal judgments, (3) vivifying of specially interest- 
ing judgments by imaginative visualization. ^*^ Supra, n. 323. 366 infra, on the Th^wtet. 

3*') Theostet., 161 E, ikiyx^i-f Tat iWriXmv i/iavrao-iar 7€ «ai ^^T Supra, n. 301. 

Sola?. 3^NeuePlat.Forsch.,p,b2. 369209 3. 

Paul Shorey 49 

including judgment, and Bo^a may always be used either of the belief that accompa- 
nies aicTdr}(n<;, or of the operation of the mind as opposed to sensation. 

Campbell thinks the rejection in Politicus, 281 C D, of KaXKvTTqv koX fieyia-TTjv 
iraamv as a satisfactory definition is an advance on Thecetet., 207 D, where the sun is 
defined as the brightest luminary, etc. But the point is simply that made "already" 
against Gorgias's fieyiara t&v avOpeoireieov irpayfidTcov as a definition of the matter of 
rhetoric.^ Again, Campbell thinks the mention of Bo^av and <f>avTa(7iav in Sophist, 
260 E, as distinct faculties implies an advance on the Thecetetus. But the Thecetetus 
does not identify the words by using them once or twice as virtual synonyms. The 
Sophist, 264 A, temporarily distinguishes (jiavraaia as a judgment present to the mind, 
St' ato-^Tjff-em?,**' while Bo^a is a judgment, iv ^^vyri Kara Bidvoiau .... fier^ aiyri<;. 
But to press this would prove too much by distinguishing the Sophist from the late 
Philebus also. 

Lastly, Lutoslawski argues °*^ that the Phcedrus and Thecetetus are later than the 
Eepuhlic, because they familiarly employ BvvaiM<; in a sense first explained in Republic, 
477 C. He overlooks Protag., 330 A, and the five occurrences of the word in Char- 
mides, 168, in a passage fully as metaphysical and abstract as that cited from the 
Republic. Indeed, the case cited from the Phcedrus, 246 D, irTepov Bvvani<;, is a mere 
periphrasis like ^ re toO vrepov <f>va-i,<s, 248 C, and of the two cases from the Thecetetus, 
158 E closely resembles the Charmides, using the word in the vague general sense of 
power or potentiality, and 185 C, ^ ye Bia ttJ? yXdnTri^ BvvafiK, uses it of the senses, as 
do the Charmides, 168 D (uko^, o-^ts), the Republic, 477 C (oy{nv Kal ukoi^v), and the 
Protagoras, 330 A (o(/)^aX/w's &Ta). Of equal value are the developments which 
Lutoslawski finds in the use of BioKeKTiKr], <f>i\oa-0(f>ia p,e0oBo<;, r) tSiv Xoycov Te^vrj, etc."' 


The dialogues were composed in some order, and a study of their parallels, coinci- 
dences, or variations in thought will often seem to indicate the plausible, possibly the 
real, historic sequence. That is not the purpose of this paper. I wish to show (1) 
that our conception of Plato's philosophy is not appreciably affected by placing the 
dialectical dialogues — the Sophist, Politicus, Philebus, and possibly the Parmenides 
and Thecetetus — after, rather than before, the Republic; (2) that the evidence is at 
present insufficient to date the dialogues of the "earlier" and "middle" Platonism, 
and that, again, from the point of view of the interpretation of the content, it does not 
greatly matter. The chief value of such negative results is that the way to them lies 
through a further positive interpretation of Plato's true meanings. 

There are certain perennial puzzles of language or thought that present them- 

360 Gors., 451 DE. " In earlier works Plato used the term soul as free from 

361C/. Themtet.,mC; s«pm, p. 48, u. 350. every ambiguity. Here we see already a trace of doubts 

about the existence of the soul. He might as well say 

362 Pp. 331, 396. that the existence of the soul is called in question by Crito, 

363 Cf. the statement, p. 373, d, propos of the innocent 48 A, Uklvo 5 Tt ttot' co-tI, etc., or by Symp,, 218 A, riji' KapSiav 
phrase, Thecetet, 184 C, «t€ »^uxV "« o ti Bel Ka\eiv that: ij \l/vxhv yap ri 6 ti Ul bvofj-Mai. 

50 The Unity op Plato's Thought 

selves to Plato in three forms: as mere eristic sophisms; as hindrances to a sound 
logical method; as serious problems of epistemology and metaphysics. They may be 
roughly enumerated as the problem of Being and not-Being, or the true nature of 
predication and negation; the antithesis in thought and things of the one and the 
many, the whole and the part, permanency and change, rest and motion ; the nature 
and possibility of real knowledge, and the meaning of consciousness of self. They are 
all directly or indirectly involved in the theory of ideas, but we may also study them 
in the group of dialogues in which they are most prominent. 

The Euthydemus presents a broad burlesque of all the chief sophisms of eristic. 
The Parmenides systematically exposes all the antinomies concerning the one and the 
many, the whole and the part, rest and motion, that can be deduced from the abuse of 
the ambiguity of the copula. The Thecetetus covers with persiflage the forms of 
eristic associated with one-sided theories of knowledge, especially materialism and 
extreme Heracliteanism, and makes a serious effort to solve the epistemological prob- 
lem. Here perhaps, and here only, does the Socratic avowal of perplexity express 
Plato's own state of mind. The Sophist makes explicit the lessons implied in the 
Parmenides and Thecetetus, and finally disposes of fourth-century eristic so far as it 
affects the presuppositions of practical logic and sound method. The Politicus applies 
the method of the Sophist to the definition of the true statesman, reaffirming from a 
different point of view, and perhaps with less confidence in the ideal, the chief doc- 
trines of the Republic. The Philebus restates the true logical method that emerges 
from eristic or metaphysical debate and applies it to the ethical problem of the 
summum bonum. 

We will begin with the Sophist, which contains the fullest exposition of method 
and the most explicit analysis of the fundamental eristic sophism. For our purpose 
there are three topics ; (1) the method of definition by dichotomy ; (2) the problem of 
Being and not-Being; (3) the logical and grammatical analysis of the sentence. 

1. The formal dichotomies of the Sophist and Politicus lend these dialogues a 
very un-Platonic aspect. They may be said to be characteristic of Plato's "later" 
style, so far as this can be true of a feature that is less prominent in the Laws than it 
is in the Gorgias or Phcedrus. Their significance for Plato's later thought is very 
slight. To understand this we must distinguish the elaboration of a definition by 
successive dichotomies from the more general logical use of distinction, division, and 
classification. Aristotle is at great pains to prove that the method of dichotomy 
assumes and does not establish the definition.^ His criticism may have been needed 
against literal -minded pupils of the Academy. Plato obviously is amusing himself 
by playing with the method.'** He clearly recognizes that formally correct dichoto- 
mies may lead to half-a-dozen definitions of the same object.^" All depends upon the 
tact with which the original "one," the concept to be divided, is chosen,**' and the 

suj_nal.Pr.,31; Anal, Post., II, 5; Part. An., 1,2 B. 3&7 md., 232 B, aWi' axoAi/Sco/iev tv irpuroi' tUv irepi rot 

•'65 See BONITZ, pp. 180 ff. croi/mrTTjr €ipi]fi€viav. ev ydf> Tt jiioi /laAio-Ta KaT€<ttdvTj avTot 

■MiiSoph., 231. i^vioi'. 

Paul Shobey 


insight that selects at each turn*^ the most significant principle of subdivision. The 
process of dichotomy is only a mechanical aid to exhaustive search and the discovery 
of all relevant distinctions.'*' The elaboration of it as a method of definition in the 
Sophist and Politicus is a mere episode. It is not followed up in the Philebus, 
Timceus, or Laws, and is therefore of no importance for Plato's "later" thought. 

A very different thing is the broader use of the method for the avoidance of 
eristic equivocation and the correction of hasty generalization or inarticulate empiri- 
cism. To distinguish and divide for these purposes is still the only way of clear 
thought and accurate speech, and Plato's insistence upon it as the one principle of 
logical salvation is worthy of the keenest dialectician that ever lived. But in this 
larger use the method kut' etBr] refiveiv is by no means confined to the Sophist and 
Politicus. There are hints of it in the Symposium?^" The Gorgias employs it with 
some ostentation.'" It is found in the Phcsdo,^''^ the Cratylus,^^^ and the Thecetetus.'^* 
Its terminology and use are familiar to the Bepublic."^ Most explicit is the Phcedrus, 
which not only makes an ostentatious display of divisions and subdivisions/™ but 
describes the entire procedure of true method in language that closely resembles the 
summing up of the whole matter found in the Philebus.^^^ But side by side with 

368 Note KwriSelv, Soph., 232 A; Poiit., 266 E, etc. 

369 The imagery of the Sophist and Politicus implies 
this throughout. Cf. Soph., 235 C ; Polit., 258 C, 260 E, 262 A, 
TO ^TjTovfiei'oi' €v StTrXaai'otffi Ta vvv iv rots 7j^i^i<re<rt.v et9 TOTf 
iroi^ffet ^rjTfttTdai; Soph., 229 D, ei aTO/ioi' 1^677 effTt nav, yj Ttva 
exov ht.aipeiTLV a^Lav 6iTtavVfj.ia^i Phoedr., 227 B, Kar' tlSrf tiexpi 
Tou aT/l^TOU TifjLVeivs Phileb., 13, 14 B, Tliv Toivvf StOM^opoTijTa; 


^T^Symp., 205BCD, a^eAovTcs . . . . rt €1605 . , . , ev 
fLopiov anpopLadev to ?repl , . , , ol ^ev aWj] rpeirdjuefoi .... 01 
Se Kara ev rt etSos tdi/res. Cf. Polit., 262 D, to fief .... (i)S ev 
.... a^aipoucTe? .... Kal -yecos €V avTO elvai. Soph., 222 A, 
iicTpe'irecrSoi'; PoiJ*.,258C; Tim., 60B,yivOi iic nivTiai' iijiopiaSivi 
SopA.,229C, 257 0,268 D. 

371 454 £, Svo cISi) eSifief. The two eiSi) are denoted, as in 
the Sophist, by adjectives in -ko^, 455 A, frequent also in 
pp. 464, 465. Soorates's humorous definition of rhetoric, 
pp. 462 ff., is in the vein of the Sophist, It starts from the 
alternative art (science) or not-art, 462 BC, like Soph., 
219 A; Polit, 258 B. It is found to be a branch of the 
pseudo-art KoAoiceuTtK^, which is divided rirpaxa, corre- 
sponding to a four-fold division of art obtained by two 
successive sub-divisions. Similarly Sophistic is finally 
found to be a part, indpiov. Soph., 268 D, of the quadripartite 

372 79 A, OStfifv .... &v6 eISt], etc.; 90 B, avev t^? irepl tous 
XoyoviTexVTjir 75 D, oI« e7ri(rijipayt^6p.£9arovTOO eari. Cf. Phileb., 
26 D; Polit., 258 C. 

373 In 424 CD, the division of letters KUTi eiSij and the 
subdivision of these eUr) is the method of Philebus, 18 B C, 
We are further required to examine the things to be named 
by letters and see ei iv avToIff tvetrTiv eiSij, and then apply one 
set of eiSi) to the other, precisely as in Phcedrus, 277 B. 

374 147 D, iireiSi) aiTEtpot TO irA^flos .... ivKkafieZv eiv iv 
(fif. Phileb., 18 B, OTac tis to aneipov avayKa<r9jj TrpStTov Kap-fii- 
v^iv, etc.) ; 147 E, tov aptOfiov iravra Sixa dieKa^op-ev, etc. 

375 397 B, ra Svo elSi); 440 E, 445 C. In 454 A, eristic arises 
fita rb jLti) SvvatrOai kot eiSi) fiiatpoujucf ot to X^ydp-tvov kiri.ftKOTtilv, 
precisely as in Polit., 285 A. Cf. Phileb., 17 A; SopA., 253D. 

Again, cf. Hep., 470 B, 5vo TavTa ra hvoiiara .... ofTa ejrl 
Svoiv TLVolv Bia^opaiy; 532 E, Kara TTola Sij eiSij SUanjKev; with 

which cf. 504A; Phileb., 23D, and Polit, 260C, Tr,v , . . . 
Texi^f .... ^eaTeoi* ei irjj BtitTTrjKev with context. Compare 

further 544 C D, ^ Tts fcal ev elSet SLa(f>av€Z TiVL KtiTai. with 
PoZif., 285 B, Sitw^opas .... inroaainep ef eiSecrt KcifTat; 580 D, 
SL^pi)Tat, Kara rpio elST], oi/Tto (eat ypvx^ .... Tpixs* 

376 244 E, 253 C, 210 B, 271 D. 

377 It is often affirmed (Jowett, Natorp, Jackson, Bury, 
etc.) that the method of the Philebus, Politicus, and Sophist 
is more advanced than that of the PhcedrtLS, in which " the 
complementary methods of generalization and division are 
applied merely to the discovery of Socratic definitions 
with a view to consistency in the use of debatable terms." 
Well, the. subject of the Phcedrus being the necessity of 
basing rhetoric upon definitions and dialectic, that point 
is naturally emphasized there (265 D, 'iv iKatnov opiCop-evot 

SrjtiOV iroijj, irepl 00 av del SlSatrKftv edf^j)). But all theories 

of a sharp distinction between the method of the Phcedrus 
and that of the "later" dialogues will only injure the 
scholarship of their propounders. The Phcedrus requires 

Tijv op.oiOTi)Ta Ttav ovTtav SiiiSivai (262 A; c/. Soph., 231 A, SeZ 
TTaVTUiV paKiaTa irepi Tas OfiOiOTtjTai iroieZadau Tljf i^vXaKiqv; 

Phileb., 13AB). To do this we must know o Jo-tiv tKoxnov 
tUv ovTioi/ (262 B). The method is twice described (265, 266, 
and 270 D). We must first reduce to unity Ta iroAAaxn Sieo^- 
7ia.pp.iva. (265 D; cf. Phileb., 16 D, aei piav iUav irepl irafToi 
endffTOTe flejiei-ovt itfTeZv; cf. 26 D). This Unity we are to 
divide icaT ap9pa n Ti<t>vKt (265 E; cf. Po!i«., 262, and with 
xaTay^vVai cf. Polit, 287 C, 265 D, icaTaflpaueii') and subdivide 
(266 A, TtpLvaiv oiiK eirafij/te), distinguishing and following up 
separately the right- and left-hand paths (266 A, Sejii .... 

dpUTTcpa; cf. Soph., 264 E, TTopeiieirSai icaTa Toiirl Se|la del (lepos 

TOU T)jr|W>'TO!), till the object of our search and of our praise 

52 The Unity of Plato's Thought 

what seems to us the purely logical treatment of the ideas as conceptual genera and 
species, the Phcedrus pictures the prenatal vision of them; the Republic announces 
the most naive realism with regard to any and every universal; and the Timceus sol- 
emnly reaffirms their objectivity.'" In the face of these facts, it is impossible to 
maintain that the dichotomies of the Sophist are evidence of a later doctrine in which 
the transcendental or naively realistic idea is discarded for the genera and species of 
conceptual logic. The emphasis and center of interest may shift from dialogue to 
dialogue — the doctrine remains the same. 

But the opposition between the two points of view cannot be denied or disguised. 
The noumenal idea is one. But not only as reflected in things, butjsjubdiyidedJby- 
logic, it is many. By a natural and inevitable metaphor both Plato and Aristotle 
speak of particulars and lower species as parts of the higher conceptual whole to which 
they are subordinated. By the theory of ideas, as we have said, each of these parts, 
every subordinate concept, is an idea, not only the summum genus and the lowest 
species, as animal and dog, but the intermediate groups, mammal and quadruped, etc. 
The Aristotelian objection that the one dog will thus embody a whole series of ideas 
we have dismissed with the metaphysics of the subject. The relation of the particular 
to the idea is a mystery. And once we have accepted the metaphors "presence," 
"participation," "pattern," a number of ideas can be reflected by or present in one 
thing as easily as can one idea. 

But the elaboration of logical and scientific classification brings up the difficulty 
in a new and more specific form less easily evaded. For the theory of ideas any and 
every subordinate group apprehended as a conceptual unit by the mind is an idea."' 
For sound logical and scientific classification only true genera and species are ideas — 
not necessarily "true species" in the sense of the mod'fern naturalist, but in the sense 
of the Platonic logic; that is, classes and groups based on significant and relevant 
distinctions. From the one point of view we expect every part to be an idea ; from the 
other, Plato explicitly warns us against mistaking for true ideas what are mere frag- 
ments or parts. ^^ His embarrassment shows that he felt the difficulty. Sound 

and blame is found (266A; cf. Soph., 235C, iwdKoXovOelv etc.; PoUt., 285A, etc.; Laws, 894AA, 963D, 965C. Each 

aiiTtZ SuLLftovvTa? , . . . emairep av At)(^0^). He who can thus dialogue brings out some aspect of it less emphasized in 

look 615 If Kol ewl iToXXa. is a dialectician {266 B C ; cf. Par- the others. We cannot expect Plato to repeat himself 

men., 132 A, Mta Tiff laiot SoKeZlSea elvai eirl iravra iSovTLf Soph., verbatim. But these variations have little or no signifi- 

235 C, TTjv Ttiiv ovTai 8vvafj.dviav ixerUvat, Koff iKoxna. Te Kal ejrt cance for the evolution of his thought. 
iravTa ^e'^oSoy) . Again, looking at it from the point of view 375 ^y^^y^^ p 35 u 238 * d S7 u 256 

of science rather than of rhetoric and dialectic (270), the 
object of investigation is either simple or manifold. If ^'">Bep., 596 A., 479 D; Soph., 225 C, ravra iiriov ^iv elSo?, 

it has many cISt), we must enumerate them (270 D, raCra ipi9- «"•«'»•«/> aurb SUyvmntv is irepov iv 6 Advos, irap eirai^v^ia? 

liTimJ-ifcvi; Cf. Phileb., 16 D, V" «" tis tov ipLBftov avroi o"" viv v^' ii^liv -nixtlv ofior. PAtJeft., 18 C D, the Stir^ds of 

Trai-Ta Kartfj) tov ij.tTa(v toO ijreipov re «al tov enjs), and treat association in our minds makes a unity, and hence an idea 

each subordinate ev (cf. Phileb., 16 D, Kat rStv iv iiteivtav sKa- ^^ ypa-fifiariKj]. 

tmv iraXii/ ilo-avToij) as we do the original unity — i. e., study 380 PoUt., 287 C, implied " already " in Phoedr., 265 E ; 

its potentialities (5uVa/xi9, active or passive; cf. Soph., c/. Polii., 262 B.aAAa to /Ac'pos ajuta elSos exeroj. We are more 

247 DE) in relation to other things. Rhetoric is a special likely to "meet with ideas" if we bisect the universal 

psychological application of this general scientific method. (fiKroTojieiK) and proceed by successive dichotomies, than if 

It is one method which is described in Phxzdr., 265, 266, we attempt to separate the ultimate species at once. Cf. 

270D; Phileb.,Vo-l%; Cratyl.,i2AC; Soph., 226C, 2350,253, the insistence on ra >i.eo-a in PMJeh., 17 A. 

Paul Shoeet 53 

method required him to emphasize the distinction. But he was quite unable to define 
its nature.^' The nominalistic logic of the modern "flowing philosophy" of evolution 
would meet the problem by making both "true species," and the tentative species of 
imperfect or erroneous generalization alike relative to the purposes of man — working 
hypotheses, instruments of greater or less precision and range, employed by thought in 
the effort to shape in its own image or check for its own ends the ever-flowing stream 
of change. 

Plato would have preferred mystery and self-contradiction to this as an ultimate 
philosophy. But his logical practice approaches nearer to it than does any interme- 
diate compromise of common-sense from Aristotle to the nineteenth century. Psycho- 
logically and ontologically all universals, as opposed to sensations and images, are 
equally noumenal ideas, whether language provides a name for them or not.'*^ In 
logical and scientific practice the only ideas worth recognizing, whether named or not, 
are those that embody significant distinctions relevant to the purpose in hand.^^ The 
recognition that words are mere counters'** and do not always stand for (relevant) 
ideas ''^ is an apparent, but not real, contradiction of the abbreviated formula of the 
Republic that we assume an idea for every word.*'* Similarly, as we have already 
seen, the occasional and inevitable use of conceptual language is no derogation from 
Plato's philosophic realism.'*' Practical logic and psychology must treat ideas as con- 
cepts, whatever else or more they may be. 

2. The puzzle that false speech and erring opinion are impossible because we 
cannot say or opine that which is not, is nothing, must be translated into Greek to 
win even a semblance of seriousness. To appreciate Plato's achievement in disposing 
of it forever we must have studied it in the poem of Parmenides and in the eristic 
of the fourth century.''* Our problem here is the seeming contradiction between 
the Republic and the Sophist. The Republic distinctly avers that it is impossible 
even to opine that which is not — thus apparently yielding to the fallacy."' The 
admirable analysis of the Parmenides and the Sophist explains it by pointing out 
that is, in its double function of copula and substantive verb, is ambiguous,"" and 
that this ambiguity extends to the convenient Greek idiomatical use of the parti- 

381 Polit.t 263 A B, to distinguish genus (or species) and 386 59$ A. The common name of n-oAAa does imply a con- 
part would require a long discussion. He can only say ceptual ev, which implies an idea, though it may not be 
that, while every species is a part, every part is not a relevant or worth while (afioi- eiru>'U|ni'a!) for the classifica- 
species (elfios). tion or purpose in hand. 

382 Supra, p. 37, n. 250. 387 E. g., Phcedr., 263 D E, iivayKairev ^jiSs inoXafic'iv .... 

383J2ep., 445 0,544 AD, 17 riva aAXijv ex«« iSiav iroXiTeitts, gy yt jSiv ovTutv, etc.; Folit.^ 258 C, 5vo cIStj BiavorjQ^vai 7r}V 

^Tts Kal €V eiSei Stai^avet Ttvl KEiral; Tim.t 83 C, €ts n-o\Aa /itt- tpyx^iv ijtiSiV jroi^o-at; Phileb,^ 18 C D, 23 E, vo^ffat, n-fl jroTe ^v 

KoX avoiioia pAejrctv, opal' 6« €v avToU ev yeVoff kvhv a^iov CTTiuru- aVTtov fv Kal iroXXi eKarepov, See SUpra^ p. 39, n. 264. 
,.;«; Soph., 229 D, y.vajxo" a.a.>e^.v a^«. e^..vv^i«; 223 A, ^^j g^^ ^ j ^^j Xjj ^^ g ^^^ ^^, ^^ 

225 C, 267 D, names for ideas often fail because the ancients „ „ 
were neglectful of t^s tS>v yevuv Kar elSij 5iatpe(rccu9. Polit,, 

ZeOE, aviiwiiov . . . . ivoita irepov airroi! TTapoiX'^priirai'Tti ee(reai. 389 478 B. Cf. Parmen., 132 B C, 142 A, 164 A, 166 A; 

Tiya; 261 E, to /iij airovSaiciv ejrl Toil oi/d/iao-i, 263 C. Thecetet., 167 A, 188 D. 

384 "Already," Charm., 163D; Polit,, 261E; Thecetet,, 390 pormen., 142 C, "vf ie ovk aiirri i<rriv i u7ro9eini, ei if 

168 B, 184 C J Soph., 218 C ; Laws, 627 D, and passim. Iv . . . . oAA' ei Iv itrrip ; 163 C, to Si /xij ia-nv .... ipi iirj ti (!\Ko 

3S5Soph., 217 A; Polit., 263 C, oti iratri tovtov en-oi'o/jia^cti' (rqfj.aivet r] ovaiai airovtriav; 162 A B, with my interpretation, 

iaxK ovojia; Sep., 454 A. A. J. P., Vol. XII, pp. 349 S. ; Sophist., 256 D E £E. ; Tim., 38 B. 

54 The Unity of Plato's Thought 

ciple — ov and /ir] 6v, ovja and fii) ovra; that firj ov is not nonentity, but otherness; 
not nothing, but some other thing. ^' If we can show that other dialogues, pre- 
sumably earlier than, or contemporary with, the Republic, ridicule the fallacy, or imply 
the answer to it given in the Sophist, we have established a prima facie presumption 
for an interpretation of the Republic that will remove the contradiction.™^ This is 
the case. In the Euthydemus the jxrf ov puzzle is one of the stock fallacies of the 
eristics. To desire to make Kleinias wise is to wish to make him other than he is, 
what he is not — not to be. The suggestion enrages Ctesippus, but Socrates bids him 
jxr) ovo/iaTi hia^epecrOai.^^ And when the quibble is further invoked in support of the 
paradox that y^evhrj Xeyeiv and y}revBrj<; Bo^a are impossible, since we cannot opine or 
say what is not, Socrates observes that this opinion refutes itself as well as all others, 
and declines to take it seriously.^ In the Cratylus Cratylus argues by a fallacy, else- 
where exemplified in Plato, ™^ that a bad law is no law, an unapt name is no name, and 
a false statement is no statement, because it is to nfj to, ovra Xe'yetv.^ Socrates dryly 
observes that this thesis, though it has many supporters, is too subtle for him,*" and 
then proceeds to offer a perfectly sufficient practical explanation of the difficulty by 
means of an illustration analogous to the image employed in the Thecetetus^ to 
account for certain forms of mental confusion. As you may wrongly assign A's pic- 
ture to B and B's to A, so in the use of terms it is possible to apply X to A and Y to B 
when the opposite distribution would be correct, and, in the case of words, true.^ 
This explanation Cratylus is urged to accept in order to avoid (eristic) debate, Xva fii] 
Haxa>ij''e0a iv rots Xo'7oi9.*"° And when he yields, Socrates commends him on the 
ground that this is not the place to argue the question.*"' There is a further anticipation 
of the Sophist in the suggestion that those who insist on the quibble are oyjniMaOei'}.*^ 

391 It is tme that Plato nowhere states the ambignity most effective analysis of the fallacy in the form in which 
of the copula with the explicituess of Aristotle and John Greek usage presented it. Plato is, for the rest, aware of 
Stuart Mill. But the passages cited in the preceding note the distinction between contradictory and contrary op- 
prove that he understood it perfectly. Grote, in his criti- position (Symp.^ 201 E; Farmen.^ 160BC; Soph., 25TB, ovk 

cism of the Sophist, objects (1) that Plato fails to distin- dp', evavriav orai' dlrdt^atrts Ae'yijrat tnjftalvetv trvyxupijirofieBa), 

guish €<TTLv in its function of pure and simple copula ; (2) and he understands the use of e'i'at as a copula, though the 

that the (absolute) other of Being is just as meaningless religious and metaphysical associations of " Being " cause 

as absolute not-Being; (3) that negation is something dif- him to stigmatize it as " inexact " (Tim., 38 B). 

ferent from otherness, and that to define it as otherness is 392 My task would be much simplified if I could accept 

to confuse the distinction between contrary and contra- Natoep's view (Hermes, Vol. XXXV, p. 425), that the rela- 

dictory. These criticisms ignore the difference between y^e Being of the Sophist is distinctly anticipated in 

Greek and English idiom, the necessity that Plato felt of Phcedo, 79 A, «iio tlS-n rav ivrav to iiiv iparov, n «e aei««. But 

meeting the ij-v ov fallacy in its own terminology, and the ;,.tco>' is not to be pressed here, 
religious or ontological associations which half playfully, 
half seriously, he was resolved to preserve for tlvai. to tirj 

393 Euthyd., 283, 285 A. 

iv, besides its ontological meaning, can be naturally used ^^ 286 C, where, as in the Themtet, it is attributed to 

in Greek idiom as a mere category embracing all particular Protagoras with a malicious aUusion to iA^Seia. 

cases of (o) negative predication, (b) misstatement. Any 395 429 B ; c/. Hipp, ma/or, 284 E; J/jtios, 314 D ff. 

particular /i-ij ov is something other than the corresponding 396 429 D. 

ov ; and, generalizing, Plato may say that ni ov is the other „, , . . , ^ „ , „„ „ 

' , . .^, ^. , . i, ... • it 11. . I, 1 1 397,oni((oTe))Oi (levo Aoyos 1, itaT e(ie, etc.: C/. SopA., 239B. 

of the ov without implying that it IS the other of absolute aogioiP v^., v.uviy«., ioox.. 

Being. For the same reason, in explaining the nature of 

error and misstatement, he is justified in substituting for '" *30 D, eirl Se Tolt i^ojiao-i Trpbs riZ op6r,v Kal iAi,9^. 

the general category jii) ov a concrete (affirmative) mis- *™ 430 D. 401 431 a. 

statement, " Theaetetus flies." It all sounds crude enough, *oa 433 A, fidf ai/iev avT^ rjj aAijdei^ ovtw n-ws ikuKv^ivat, oi/iiot- 

if we think it only through English idiom. But it was the Tepoi' toC Scoi-toi. Cf. Soph., 251 B, 259 D. 

Paul Shoeey 55 

It is obvious (1) that the fallacy is none to Plato ; (2) that he feels himself able to 
carry the analysis farther; (3) that he does not do so because he wishes to write the 
Cratylus, not the Sophist. 

In the Theoetetus the matter is somewhat more complicated. As we shall show 
more fully below, the object of the Thecetetus is not to refute or analyze the logical 
fallacy that false opinion is impossible, but to explain the psychological nature of 
error, and with it of cognition: ti ttot' earl tovto to ird6o<; Trap' ^fiiv ical riva rpoTrov 
iyyiyv6fievov.*°^ For this the /Mr) 6v quibble would have been wholly unfruitful. But it 
could not be altogether ignored. Hence it is perfunctorily dismissed in a page with 
the admission that the method of elvat and iirj etvai offers no explanation of error, since 
6 Bo^d^mv ev ti 8of afet, and o firjSev So^d^cov to irapdirav oi/Se fio^a^et.™ We are thus left 
free to pursue the psychological analysis kuto, to elSevai koX p^r). But it is absurd to 
suppose that Socrates is really baffled in the Thecetetus by a fallacy at which he 
laughs in the Euthydemus and Cratylus. And his real opinion of it is sufficiently 
indicated by his attribution of it to Protagoras in this very dialogue."^ 

The final analysis of the fallacy in the Sophist is introduced and accompanied by 
persiflage in the manner of the Euthydemus and Cratylus, and by hints that it is a 
mere eristic puzzle.*" The final common-sense formula that true speech and opinion 
represent to, ovtu oj? e^ei or w? eVrt is not new.*"' It evades the psychological prob- 
lems of the Thecetetus, and it is reached by arguments purely logical and practical. 
If we do not admit that prj 6v normally means otherness rather than non-existence, we 
shall make all rational speech and thought impossible.*™ The absolute 6v (and pr} 6v) 
of the Parmenides to which no intelligible predicates attach is reserved for ontology 
and mysticism.*"' But iv rot? irap fjplv X6yoi<} (251 D) we must accept a doctrine of 
mixed and relative Being and not-Being.*'" 

The result of the inquiry is that, if Plato in the Republic falls into this fallacy, 
the Republic must be earlier and less mature, not only than the Sophist, but than the 
Euthydemus and the Cratylus. But Plato does not yield to the fallacy in the Repub- 
lic. He merely varies his terminology to suit his theme. He needs the transcendental 
absolute Being for the world of ideas as opposed to the world of sense, for the sym- 
bolism of the idea of Good, the image of the sun, the cave, and the conversion from 
the shadows to the realities. It would have been singularly tactless to preface these 
passages with an explanation that ov, like pf) 6v, is a relative term, and that all ovtw 
with which human logic can deal are likewise p-q ovtu. There is no occasion for the 
ovra and p^ oma of practical logic here. Absolute not-Being is consigned to total 

^^ 187 D. ^* 188, 189 A. 407 263 B, Aeyei fie a.VTu>v o fikv aAi)di}s tol ovra ws eo-Tt irepl trov. 

*05 In Socrates 's ironical defense of uUra-Protagorean- Qf.Cratyl.,dS5'B,osavTa.6vTa\€yT[ibi^e<TTiva.KTij9i^i; Euthydem.^ 

ism, 167 A, oure yap Ta fiij ovTa Swarbv 'fio£a<ral, ouTe oAAa irap* 284 C, aWa Ta oi'Ta fikv rpoTTOV Tt»'a Aeyet, ov /xefrot ws ye ex^t. 

i av Trao-xn. Cf. Cratyl., 286 C. log 238 C, 239 B, 249 B C, 252 C, 259 A, iSivSv cip^Kaiitv eUai 

m 236 E, ei/ai/TioAoyi? (i>| <rvWx«ir9al, etc. ; 237 B C ; 239 B, ^0^11 ov, ^ Treiffi™ tw lis ou /caAit Aeyo(ie>' eAeyJas r, mW' "'P 

i^i . . . . iroAai KoX Ta vvv rimu^ivov av tvpoi irepl rhv To5 /<,), -^ iSui-aTB, KtKTiav ual ««V(j. icaOairep iifjiclt, etc., 260 A. 

Si-TO! eAeyxoi', etc.; c/. 242A, 243AB, 252C. Note also the laovanv t to 

close parallelism of this part of the Sophitt with the inten- 258 E ; c/. snpra, p. 39. 

tional fallacies of the Ponncntdes, infra, pp. 58, 59. 4io 251 A, 254 C D, 259 A B. 

56 The Unity of Plato's Thought 

ignorance as it is in the Sophist.*" Pure Being is reserved for the ideas, as it is in 
the Timceus, which was written at a time when the results of the Sophist were cer- 
tainly familiar to Plato. Its antithesis, the world of phenomena, is described as tum- 
bling about between Being and not-Being — as a mixture of the two; the things of 
sense are always changing — they are and are not.*" It is not necessary to dash the 
spirit of mystic contemplation and enthusiasm by the reminder that the ideas them- 
selves, when drawn down into the process of human thought, move to and fro and 
partake of both Being and not-Being."' We are concerned here only with the broad 
contrast between the two worlds. To say that the objects of sense and the notions of 
the vulgar tumble about between Being and not-Being, is merely another way of saying 
that they belong to the domain of the mixed or relative Being and not-Being described 
in the Sophist."* Only a deplorably matter-of-fact criticism can find in this adapta- 
tion of the terminology to the immediate literary purpose a concession to a fallacy 
ridiculed throughout the dialogues. And the arguments that would prove the results of 
the Sophist unknown to the author of the Republic would apply almost equally to the 
Timceus; for there, too, Plato calmly reinstates the absolute ov which the Sophist 
banishes from human speech as no less contradictory than the absolute /xi) ov, and 
treats as an inaccuracy the expression to /xtj- ov jxt) ov ehai, the practical necessity of 
which the Sophist demonstrates."^ Yet the treatment of the "same" and the "other" 
in the ■^^vyp'^ovia (35) proves that the analysis of the Sophist was familiar to the 
author of the Timceus. 

3. The explicit discrimination of ovo/Mara as names of agents and of pij/iara as 
names of actions is peculiar to Sophist, 262. So the special definition of Btdvoia is 
confined to the Republic,*^'' and nearly every dialogue employs some definition or 
distinction which Plato does not happen to need again. Even if we concede that this 
greater explicitness of grammatical and logical analysis marks the Sophist as late, its 
significance for the development of Plato's thought is slight. It is not repeated in 
the Politicus or Laws,*" and it is virtually anticipated in the Cratylus, where it is 
twice said that Xoyo'; is composed of prifiaTa and ovofiaTa.*^^ It is barely possible, but 
not necessary, to take prifiara here in the sense of "expression" or "phrase." Even 
then it must include the verb. For ovo/jia is plainly used in the sense of "name" or 
"noun." Lutoslawski's argument"' that "it would be unjustifiable to apply to the 
Cratylus a definition given only in the Sophist," obviously begs the question. The 
expression (425 A), Kal cruWo/Sa? av crvvTi0evTe<i ef &v rd re ovo/iara kuI to, prjfiara 
avvTidevrai, seems to put ovojxara and pi^fiara on the same plane and is unfavorable to 

*ll 477 A, M ov ^TjSoifin ; 478 D E, toS niyrm M OVT05. Not 41* Cf. A. J. p., Vol. IX, p. 307. 
foreseeing modern philology, Plato did not think it neces- 415 Tim., 38 A B. 416 Supra, h. 346. 
sary to add jra^rus or (^TjSa/xJ a third time in 478 B, when he 417 Lutoslawski is mistaken in saying that pw« is used 
asks i iSvvarov kcX «ofai7ai to ^r, oy, which LUTOSLAWSKi, }„ (.j^g distinctive sense of predicate in Polit, 303 C, and 
p. 429, thinks would be unaccountable coming after the j,„^s, 838 B. In both places it means "saying," " state- 
inquiry of the Sophist. Similarly Apelt {Beitrdge). ment." 

412 479 BCD. 418 425 A, 431 C, Aoyoi yap TTOU ws iytaftat, if rovTuv ^vvdttris 

413 Though it is hinted in the iAA^Awv Koiviavta of 476 A; i<rrtv. 

cf. supra, p. 36, n. 244. 419 p. 431. 

Paul Shobey 57 

the notion of a progression from syllables to words, and from words to phrases and 
sentences. In 431 B, if fyqiiaTa means "verbs" or "predicates," we understand the 
statement that they as well as ovofiara may be falsely applied. But what is a false 
application of phrases ? And if we evade this difficulty by taking pij/iara as " sentences," 
then \6yoi must mean, not "sentences," but "discourses," and what is a false attribu- 
tion of discourses ? In fact, it would be easy to argue that the Cratylus takes for 
granted the results of the Sophist and is therefore later. Our concern is not with 
such "arguments," but merely to show that, conceding the utmost that the texts will 
bear, the difference very slightly affects the relative maturity of the thought iu the 
two dialogues.*''" 

A great deal of ink has been spilled over the Parmenides, and the profoundest 
mystical meanings have been discovered in its symmetrical antinomies.*^' To rational 
criticism nothing can be more certain than that they are in the main a logical exer- 
citation more nearly akin to the Euthydemus and the Sophist than to the Timceus, 
and that they are not meant to be taken seriously except in so far as they teach by 
indirection precisely the logic of common-sense expounded in the Sophist.*^' In 
style, however, the Parmenides presents few, if any, traces of the elaborate "late" 
manner of the Sophist,*^^ and this fact makes the identity of doctrine the more signifi- 
cant. Both the Thecetetus and the Sophist allude to a meeting between Socrates and 
Parmenides."* The method of argumentation employed is characterized in the Phcedrus 
as a kind of rhetoric, and in the Sophist as mere eristic.*^' Many passages closely 
resemble arguments and expressions which are ridiculed in the Thecetetus and Sophist, 
and which are presumably not serious here.*"^ The dialogue itself abounds in hints 

420 Cf. supra, p. 33, n. 218. The further points made by Plato at the time of Themtet., 183 E, intended to discues 

Lutoslawski are nearly all misapprehensions. He says rest and motion, but, writing the Parmenides much later, 

that the admission that philosophic teaching may be changed his mind and devoted Part I to objections to the 

given by continuous lecture, as well as by the method of ideas, and Part II to metaphysical problems still debated, 

question and answer, is first found in 217 C. But Thecetet., i2i Phaedr., 261 D, rov o!!^ 'EAeaTiico^ naAanii8i,v (Zenot) 

167 D, recognizes the same choice. The meaning of liiOoSot Xiyovn ov« laiJLiv rixiTt Hare 0aiVe<r9ai rois oitov'ovo-i tA aira 

in Soph., 227 A, is not more definite than that in Pficedr., s^„,a ,»i i^^^oia, koI Sv koX rro\ki, etc. Soph., 259. It is 

270 D, and Sep., 533 C ff., except in so far as the method of equally foolish to deny or to take seriously the antinomies 

the Sophist and Politicus lays more stress on the mere mech- (eravTicuo-eaii/) that arise from the communion of ideas and 

anism of definition by dichotomy. Cf. sv.pra, a.3n. The the relativity of oi-, (lij oj/, and flare/ioi'. C/. 259 D, ib « ravroF 

notion of logical exercise is not new here, but is found in irefiov iiroiitiiiveiv a/nj ye irji . . . . /cal to /ieya aiitKiiov koX to 

Merw, 75 A, ti-a ital ye'niTai eroi (leXeTT), etc., and is implied in i^oioi' afoftoiov .... ouTe Ti! iKcyxoi oJtos i\rj0iv6$, etc. 

Thecetet., 147 A ff. Dialectic in the Republic is as clearly guch contradictions are nothing difficult when one knows 

the science of the division of notions as it is in the Phcedrus the trick. 259 C, eire m ti xoAeiroi- KaTafevoijKm. Cf. Parmen., 

and Sophist. See 454 A, 535 B, supra, n. 365. See also on 159 A, leal irai-Ta ri eyavria iraflij oviceVi x<»''"'"S eup^ao/iev, and 

&vvaij.K, supra, p. 49; and on the ideas as souls, supra, p. 39. Socrates's congratulations to the Sophists in the Euthyde- 

*2iBuKT on "Later Platonism," Jour, of Phil., Vol. rnus on the ease with which Ctesippus picked up their 

XXIII, pp. 161 fl., gives a useful summary of recent discus- method (303 E). 

sions. *26 E. g., the quibble, Parmen., 147 D ff. (of which Alice's 

422 Cf supra, p. 54. De Plat, idearum doctrina, pp. 41 ff. ; "Jam every other day " is the only English analogue) , that 
ATP Vol IX DP 185 290 ff '^^ "other" is the "same" because the word erepov in 

Greek idiom applies to both, and the word must refer to 

423 Natoep, Archiv, Vol. XII. tj^^ g^me essence. This is parodied by Socrates in Euthy- 
Hi Thecetet., 183 E; Soph., 217 C. Either allusion might dem., 301 B, and explained in Thecetet., 190 E, eireiJ)) to p^»"i 

precede or follow the actual composition of the Parme- erepov ri iripa Kara p^c* TavToj- eo-Tii-. The extension of this 
nides. Natokp, Archiv, Vol. XII, pp. 291, 163, supposes that reasoning to the i^o/xoioTaToi' is deprecated as eristic in 

58 The Unity of Plato's Thought 

to that effect. It is recited by one whose light has gone out more completely than 
that of Heraclitus's sun, and who now is devoted to horsemanship.*" Parmenides 
himself characterizes it as a kind of intellectual gymnastics which it would be unseemly 
to practice in the presence of the uninitiated,*^' and explicitly terms it a TrpaynaTeimSrj 
■TraiBidv.*''' He chooses as his respondent the youngest interlocutor, on the ground that 
he will be least likely "TroXvirpayfioveiv — that is, to interrupt the flow of plausible ratio- 
cination by distinctions like those with which Socrates checked the stream of fallacy 
in the Euthydemus.*^ 

These are probabilities. The proof is that the fallacies are symmetrically 
deduced by a systematic abuse of the ambiguity of the copula, and that Plato gives us 
clear warning of this at each turn in the argument. The symmetry is of course not 
perfect, and there are various minor fallacies that arise from other equivocations. An 
analysis full enough to show this in detail would defeat its own object by wearying 
the reader and obscuring the main design, which is not open to debate.*'' The 
groups of contradictory conclusions deduced from the hypothesis that the One is and 
that the One is not derive almost wholly from the equivocal meaning of "is" — from 
taking "is" or "is not" to signify now the absolute uncommunicating Being or not- 
Being which the Sophist dismisses as impracticable, and now the relative Being and 
not-Being, or otherness, which the Sophist establishes as the only tenable use of the 
terms in human logic. And near the beginning of each hypothesis we are distinctly 
warned of the sense in which "is" and "is not" must be taken.*''' This is perhaps 
sufficient ; but another way of putting it will bring out the parallelism with the Sophist 
still more clearly. The eristic combated in the Sophist may be resumed in two fallacies: 
(1) The noumenal unity of the idea is incompatible with any suggestion of change, rela- 
tion, or multiplicity. The ideas will not communicate or mix. Predication is impossible. 
You cannot say, "Man is good," but only, "Man is man" and "Good is good."*'' 

Phileb., 13D. The Parmen., 148 A, infers that Kar' ovto wo Lutoslawski, p. 418, niisuDderstaiids this, saying: 

toOto airav airtun ojlioiov av elij. Now, it is precisely the func- " It is only in the Parmenides that discussion (woAvTrpoy- 

tion of deceptive rhetoric Tav Travrl ojmoiovr, Phcedr., 261 E; it.wkiv) is declared useless." 

and it is precisely this that the Sophist, 259 D, and the 431 ggg Apet.t, Beitr&ge. 

PftiJefcMS, 13 A, stigmatize as eristic. Similarly the antino- i,o/,\io7t> •■ - ■ • /o\ <i»n - r- • 

miesofwholeandpartinl37CD,144E,145E,157E,159CD, . .*'! '.".f ^'."'r ""^' " 'V .<?' ."^ ^' -"'' ^' Z\^:V! 

recall ThecBtet.^ 204, 205, and Soph,, 245. On rest and motion 

i(Tr\v 71 uTTofleo'iy €t I** ec . , , , aAA' cl tv €<rTiv . . , . ; (4) 157 C, 

of. 139B with Soph., 250C, 146A, 156E, 162B, with 255Ei '"'^' ""'' <'r'P"a.y.„a.r..acr.rov..o,riX^a.a>J.^ 

The<ztet., 181-3. In Thecetet., 180 D, the words Wa .a. ol ' • ■.■ ;,<"^*™ f^l' ^^^^'^O ""7" Jf""'!, "^r " ^^ ™7^'^'»'- 

..-,., » » . , ^ X*^P^5 0€ TdAAa Tov €1/05 €lvai; (6) 160 C, on ereaov rt Aevot to 

axuTOTOuot .... ffauffwi'Tai nAi0t&>5 oioiMevot Ta iiei' taTavat, Ta >-« .,... .-* , 

. „ - . ,. T.1 i , 1 • • * ii. Ml? o**, orav tmn tv ci /atj effTt, Kai iiTU€y o Aevet (cf. Soph., 

Sj (cii-eiffSai Tioi' ovTii)!', show Plato's real opinion of these oi.-,t>\ ■«.n-c t - 7- • •- ■ T, \ j.'^^±--., 

absolute antinomies; c/.SppA., 249 CD. For the negation „.' .,,' ^~ ' ^ ... '^ . . ■ ■ • • i^ \ 

. ,, . ^ IT -Li J- i ^ 1JO A ici -D n 1. «jo ^ iroAAbtv oufiei' KWAvet. From this OVffiaff ucrevetl' and then 

of all intelligible predicates cf. 142A, 164B Soph., 248C; , , . j j j t ,n^ ,00,-. ^. t- . - 

„, " _. ,_„v, T I ii D J Tts eii'ou MOV are deduced; confro (7) 163 C, Ti>i«/ii)«(7Tii/ ... . 

Tfterefef., 157 B. In general the Parmenides exemplmes , , .,, ..... , , / . ,, 

. ,,,«,- ..i mew ' c ox *P* ^^ T' ^^o cnj^ati-et jj ovo-iac aTrovo'iai'; (cf. Ae., Met, 

what the Sop/wet terms, 245 E, T0U5 .... «i<ucpiPo\o70vf«i'ovt ioo4a 151 y.yi . ^.i^ , ■a^'^,-, 

OVTOS Te Trepl Kai fi^. 

«8 251B, 259E, 251 C; 2Tie<E«e(., 201E-202A. The dBSiv 

*"128 C. ifiAol, 248 A (cf. 246 B, 248 E), represent not so much a par- 

«8 135 D, 136 D E. The Euthydemus hints that listening '""^" ='=^°°! «? % generalized tendency of thought. They 

to eristic m^y be a useful discipline. This is the meaning ".'^^/'"''t.f'f 1"^ Platonists or E eatics who introduce 

-..v.:_..'*,-™„f *!,.«„ ...!„,„„ SI!*!!' .„H„f9fKn „ff.„ mto logic Plato's (and Parmcnides's) poetical absolutism. 

of the intervention of the Sai/x(iKiav, 272 E, and of 305 D, often 

Plato's criticism is not a recantation of " earlier " Plato- 
nism, for their dogma in SppA., 248 C, is precisely what Plato 
*M 137 B. himself says in Tim., 38 A ; c/, supra, p. 39. 

Paul Shoeey 59 

(2) The negative "is not" denotes absolute non-existence, which is unutterable and 
unthinkable.*" Plato answers in substance: (1) We must admit the mixture of 
ideas, the seeming multiplication of one idea by communion with others, as a condi- 
tion of intelligible speech. Without it we cannot even predicate existence, identity, 
and diversity."* (2) Absolute not-Being is no more nor less a problem than absolute 
Being."" The only not-Being that finds a place in intelligible speech is otherness — 
that which is not this, but is some other thing.*" Now, in the eight or nine*^ 
hypotheses of the Parmenides these two principles are alternately and systematically 
violated and recognized — the consequences in each case being drawn out in exact 
parallelism to those indicated in the Sophist. In the absolute theses the ideas are 
taken in self -identity, in isolation, ;^a)/)i'?.*™ The one has no parts, and the exclusion 
of parts is found to shut out all predicates that imply multiplicity, space, time, or 
number.**" And since these are the forms in which Being appears,**' we cannot even 
say that it is.**' There is nei thetL-kJi&wledge nor speech of it.**' In the absolute 
negative theses fit) 6v is taken to exclude every sense of ehai, with a similar result.*** 
In the hypotheses concerned with relative Being and not-Being the reasoning is 
reversed. If we speak of unum and alia, we imply existence in some sense. The 
existent one is two (unity and existence), has parts, and so by necessary implications 
is clothed in all the predicates of space, time, and relation.*** Instead of abiding in 
isolation, the one everywhere united with essence, ova la, is divided up among the 
indefinite multiplicity of 6Wa.**° And it is explicitly affirmed that this is true of the 
most abstract and ideal unity that we can conceive.**' Similarly, starting from the 
assumption that iiij 6v (or /*»? ei*) means something, and something different,**' we 
deduce first "participation" in various predicates,**' and finally the defiant paradox of 
the Sophist that fir) 6v ian.*^ The doctrine of these relative hypotheses is that of the 
Sophist. The reasoning of the absolute hypotheses is that of the preliminary airopiai 

434 238 C- 241 A, etc. 442 141 E, ou6' apa ovTu)5 eo-Ttc oio-Te ev elvai. Damascins 

435 252 c 256 A B 259 E etc. ^"^^ iii&t Plato does not negate ir of iv, but Simplicius, 

436 250 DE, 258 E. 

Fhys., 88, 32, contradicts Mm, 
443 142 A; cf. Sqpft., 248 C ff. 

437 257 ff • 

444 163 C, 164 B, oilTtii Sij iv OVK 6v OVK exei irwff ovSafXjj. 

438 The third fri Br, to Tfiinv Xiyat^iv, 155 E, stands by 445 142 C, it 4AXo rt oT^iaiVov to eVn to5 Ir . . . . toioOtok 
itself. It is in some sort a reconciliation of the contradic- 5^ ^b Ik <n,|iiaiV£iv oUv ^ep, e'xeiv, etc. ; cf. Soph., 244 D fl. 
tions of the first two, and, by impUcation, of all. 448 144 B, cti Trivra «p. ToAAa ivra r, oio-ia v,viM-r<^i, etc. ; 

439 137 C, 139 E, To5 Se ye iv'os X"P'< iijal/ij rriv 0i!irtv To 144 C, irpbt airam apa UaKTTif Tip rij! oiiriat fiepei irpoaeaTi To 

Tavrov, 140 A, 159 B, ' Ap o5v oi x™P'" 1^^" " tt> Tav akkav, etc. f„. (y_ Soph., 245, 256 D E, 258 D E. 

Cf. Euthyd., 284 A, iv fL^v KaKelvo y eo-Ti Tii- Sxtmi-, 6 Wy" HI 144 E, oO liovov ipa. TO if ev iroXAa effTiK aWa Kal aiiTo to 

xVs f"" i^^". TftecBte*., 205 C, SioTi avro itaS' auTo «a(7Tov .^ ^^.^ ^^. .^^^^ Siavey€pir,pii>'ov; cf. USA. Republic, 525E, 

e'lT) iavvBerov, Kal ovSi to eli/a. irepi auTov 6p»i! ex"' irpoo-^epowa however, points out that thought must restore the abstract 

€itrely. Another form of this fallacy, irSi/ iirb iravriK x<»p'fe'^, unjty ^g f^gt as analysis divides it : iM' iav av KepnaTiiiji 

appears in the Protagorean doctrine: Cratyl., 385E, iSi? ^^^.^^ Utlmi iro^XairXoffioSaiy, eiXa^ovjievoi p.-f, irort ifavf to iv 

auTii- ii oidia elvai Uaarif, Thecetet., 166 C, Uiai oiir9i)<rei5 p.^ ly aWi iraWa /iopta. For the use of «epna"iu here and in 

iKiartf iip.S>v yiyvovrai. Absolutism, whether sensational or ^^^ Parmenides, cf. Soph., 258 D. 

verbal and ideal, destroys rational thought, and is refuted ^^^ ^^ ^^ ,^^ .^^^^^ ^,_^^^ Tb ^), 5^ . . . . «al o Xeyei. 

by pushing it to the extreme where this is apparent. ^^ ^^^ ^^ ^^ ^^ ^^^^ 255 A B. 

440 137 C - 142 A. Similar results follow for tIAXo from ^ jgj A, Ser ipa airh SeapLov ixtiy to5 p.ri elvai Tb elxai p.r, iv. 

taking tv \<apU and without parts 159 B- 160 A. j-^y jljg indispensable emendation of what follows, see my 

441 Tim., 52 B. note in A. J. P., Vol. XII, pp. 349 ff. 

60 The Unity of Plato's Thought 

in Sophist, 237-46, and it is well described in Thesetetus's language there (246 E) : 
(rwaTrrerai yap erepov e'f dXKov, fiei^m Kal ')(aXeK<OTepav ^epop irepl tS>v ep/irpoa-dev ae\ 
pr)6evra>v irXdvrjv. 

In view of these facts, it is idle to attempt to date the Parmenides and the 
Sophist by their philosophical content. The substantial identity of doctrine does not, 
of course, exclude many minor differences in the literary form and the secondary pur- 
poses of the two dialogues. One object of the Parmenides, for example, is to illus- 
trate exhaustively the " both and neither " of the eristic caricatured in the Euthydemus. 
The absolute hypotheses issue in blank negation. In order to make the " both and 
neither" plausible, some reasoning from the absolute point of view is introduced into 
the relative hypotheses. *'' Again, it is not easy to say how much importance Plato 
attached to the third division of the argument in which the contradictions of the first 
two hypotheses, and, by implication, of all the others, are resolved. Contradictory predi- 
cates (the "both") can be true simultaneously — they belong to different times. The 
"neither" belongs to the instantaneous moment of transition, the "sudden" which is 
outside of time altogether."^ It would be possible to read a plausible psychological 
meaning into this ingenious solution of the Zenonian problem of change.*" But it 
cannot easily be translated into the terminology of the theory of ideas. Pure Being 
admits of neither of the contradictory predicates, and the ideas as noumena are out- 
side of space and time. But the "one" which is here spoken of as out of time, and 
without predicates at the moment of transition, is apparently not the idea, but any one 
thing which may participate in the ideas. This consideration, and the fact that the 
i^aL<f>vrj'; is never mentioned again, seem to indicate that it was only a passing fancy. 

Lastly, though the main object of the dialogue is the illustration of the ambi- 
guity of the copula, and the fallacy of isolating the ideas, the one is in some passages 
a representative of the Platonic idea, and in others of the absolute Being which 
ontology and mysticism recognize even after its banishment from logic. This explains 
and partly justifies the interpretations of the neo-Platonists and that of Zeller already 
considered; but does not necessitate any serious qualification of that here proposed.*** 


The Politicus quotes the Sophist,*^^ and is closely related to the Timceus and the 
Laws. Its style and its tone of "mixed pathos and satire""' in the reluctant aban- 
donment of impracticable ideals*^' mark it as probably late. But there is nothing in 
the thought to necessitate or strongly confirm this view.*'' It cannot be shown that 
Zeller, Grote, or, more recently, Pohlman**' are led into error in the interpretation of 
the thought by their assumption that it precedes the Bepublic, and the attempts of 

«i B. ff., in 149 E-150 the denial of communion between *54 Supra, p'. 34. *55 257 A, 266 D, 284 B, 286 B. 

the ideas : oiSi rt eo-rot tr/iiitpbi' itAtji' avT^j <r/ii«pdTi)TO!. 45s 253 D, 266 B C. *5' 272 C, 301, 302. 

«2 156 B, iAX' i «f ai()>n)5 aun) i^uffi! aroiros Tts eyKa97)Tai 158 For the theory of ideas and iva/xvijo^is, cf. supra, 

fitra^v T^S KtvrjtTeia^ Kal trTatrtwq, ec xp6v<a ovSevi oiaa, p, 44, 

153 See De Plat, idearum doc, pp. 44-6. *^9 Geschichte des antiken Komraunismus. 

Paul Shobey 61 

Lutoslawski and otliers to show that the doctrine must be late are either fallacious*" 
or prove at the most that it is genuinely Platonic.*" Much of the dialogue is devoted 
to the illustration and perfection of the method of dichotomy set forth in the Sophist.*^' 
In form it is an attempt to define by this method the true statesman — to discriminate 
him sharply from other rulers and caretakers, and in particular from the politicians, 
sophists, rhetoricians, and generals, who usurp the name at Athens. '"^ 

This logical process is illustrated and its tedium relieved by a myth"* and by 
elaborate analogies from the art of weaving which also separates, purifies, and 
re-combines.*"* Kemarks are made on the necessity of thus mingling jest with earnest, 
and of employing concrete imagery or patterns to illustrate abstract thought.*** The 
charge of undue prolixity is anticipated.**' Our object is the elucidation of sound 
method and for that no briefer treatment of the theme would suffice.*** In general, 
Plato tells us, the clever men who proclaim that all things are subject to number and 
measure have neglected to observe that there are two distinct types or ideas of meas- 
urement :**° the purely relative mathematical measurement of one thing against 
another,*™ and the measurement in reference to fixed, absolute standards of the suit- 
able, the just mean or measure in every art and procedure. Long and short as terms 
of censure applied to a philosophical discussion have no meaning except in the latter 
sense. That such absolute standards exist Plato cannot delay to prove except by a 
summary form of argument employed in the same way to cut short discussion in the 
PhoBdo and Timceus."^ The proposition to be proved is indissolubly bound up with 
another proposition which the opponent can hardly reject. In this case, as surely as 
the various arts and sciences exist, so surely is the /lerpiov or absolute measure of fit- 
ness a reality. For all arts and sciences postulate it. This simple thought has often 

*60 309 C, a\ri6ri Bo^av .... tfetav ^ijjiAt ev BoLtioviff yCyveaOcu the " late " Philebus, 27 A, to SovKevov eis yevetriv aiTitt, The 

yivei does not mean that truth, etc., is " to be seen only word in an allied sense occurs in Gorgias, 519 B. It is pos- 

in divine souls," c/. supra, p. 39. In 272 C, <rvvayvpitxtv sible that it did not occur to Plato's mind in writing 

^•poirriireuK does not mean " an ideal totality of individual Pftasdo, 99B, but more probable that he deliberately pre- 

endeavoTS .... transmitted from generation to genera- ferred the periphrasis vphich is far more impressive in the 

tion." The word is used here not only for the first but for context : aA\o ij-iv ri eirri to alTiov t™ ovti., aX\o S' cKelvo ivev ot 

the last time. Campbell's citation ot Sophist, 259D, is r'o oirMv oiiK Slp iror' t'ai alnov. 
irrelevant; cf. supra, n, 439. The use of SvraiiK proves 462 See Campbell on 263 D. 

nothing ; c/. supro, p. 49. 308 C has nothing to do with the , , , , 

modem notion of building up a science by selection, «3 in 267 successive dichotomies have distinguished the 

"while useless observations and notions are rejected ;" nor statesman only as the caretaker of the biped human flock, 

with Cratyl., 438 E. The statement, 308 E, that the royal J^* remains to define his specific service to this flock, 287 B, 

art puts to death, tows /»)) Swa/ierovt Koivi^veZv, is not an '"^ "' "'^ ^ ^• 
admission of the "impossibility of proof in moral ques- «64269ff. 

tions," and in any case is virtually identical with Protag., ^^^^^^ ^^^ characteristic Platonic generalization of 

322D,Tb;.,.^,Sw^a^.™;.a.«ov5,ca.5«,5^.T«x".«T««c.^ The j,„p,„,^ „j,_ 282BC with S(vh., 226D, and "already" 

unity of universal science' is not affirmed in 258 E, or cratyl.,mBC. Cf. Phileb.,23T). 

Sophist, 257 C, except as the concept or idea (like any other 

concept) is one " already " in Sep., 438 C D. The question 

is merely: Shall our dichotomies start from the concept *68 285 D, 286. 

" science" or from some other concept as, e. g., e)«r«ipio1 

Cf. Soph., 219 A, with Gorg., 462B C. 

«6268D, 277fl. 46'283ff. 

469 283-5. The Ko/j-'poi are apparently the Pythagoreans. 

. . J r^T. t ""Trpit tt\X.)Ao, 284B. The parallel with iJep., 531A, 

461 The employment of a periphrasis in Pha)do, 99 B, for ^^^^^^^^ i„«^.Tpov.T«, seems to have been overlooked, 
the technical term avvaiTLov used m the Politicus, 281 1>, 

287 C, 281 CE, etc., and in the Timceits, 46 C, and nowhere 471284D, it apa iiyryriov ijioi'iot, etc. Phoedo, 77 A, tit ri 

else, proves nothing. A periphrasis is used tor the idea in o/wius elvai, etc. Tim., 51 D. 

62 The Unity of Plato's Thought 

been misunderstood.*" It is implied in the doctrine of ideas,*" in Plato's polemic 
against mere relativity,"* and even in the remark attributed to Prodicus in Phcedrus, 
267 B, avTo<; . . . . &v Bei Xdyav re')(vqv' helv Bk ovre ixaKpmv oire ^paxecav, aWd 
fjuerplmv. The fact that it is explicitly stated "for the first time" in the Politicus 
proves no more than does the fact that it is never stated again. Plato happened to 
formulate it only once, but it is clearly involved in Republic, 531 A, aWijXot? 
avafierpovvTei, etc. 

The myth may be profitably compared with the Timoeus, Philebus, and Laws, 
but cannot be pressed to yield developments or contradictions of doctrine. Its service 
to the argument is merely to distinguish the mythical ideal of a shepherd of the people, 
who plays providence to his flock, from the modem ruler who leaves other specialists 
to feed, clothe, and house them, and confines himself to his specific task of govern- 
ment.*" In other words, it emphasizes the demand often repeated in Plato for a 
precise definition of the specific function and service of the royal or kingly art ; and, 
as Zeller says, rejects with a touch of irony ideals drawn from a supposed state of 
nature. This ruler is further discriminated, as in the Euihydemus and Gorgias,"' 
from the pretenders or subordinate ministers who usurp his name, the rhetorician,*" 
the general,*'* the dicast.*" Lastly, his special task is defined. As implied in the 
Meno and Euthydemus, and stated in the Republic, he is to teach virtue and incul- 
cate right opinion.**" And that his teaching may be effective and the seed fall in good 
ground, he is, like the rulers of the Republic and the Laws, to control marriages and 
the propagation of the race — especially with a view to harmonizing and blending the 
oppositions of the energetic and sedate temperaments.**' 

The accompanying classification and criticism of forms of government imply no 
change of opinion unless we assume that Plato was bound to repeat himself verbatim. 
The classification of the Republic is first the ideal state governed by philosophic wis- 
dom, whether ^aa-iKeia or apiaTOKparia,^'^ and then in progressive decadence timarchy, 
oligarchy, democracy, tyranny. The Politicus apparently recognizes seven states : 
one, the right state (302 C), the only Polity deserving the name (293 C), in which the 
rulers are e-maTrjp.ove';. Six others are obtained by distinguishing the good and bad 
forms of the three types recognized in ordinary Greek usage. *** We thus get monarchy 
or royalty, and tyranny, aristocracy, and oligarchy, and democracy, lawful and law- 
less.*** The differences are due mainly to the necessity of presenting a continuous 

i'ii E.g., by Siebeok, Untersuchungen nur Phil, d. *'9 305B. 

Griechen, pp. 92 ff., who over-emphasizes the analogies with 450 309 D 

the ircpo! of the Philebus. ' 

,„_^ , . „„, . T. X ft ,. , *" 309, 310. The ieeiwWic recognizes the control of mar- 

473The ^«Tpco„ Y..e,r.5 284ABto which every artist ^age, 460, and the importance and difflcnlty ot recon- 

looks, is virtually the idea which he tries to realize, <,iii„g tj^^t^o temperaments. 503 C. It does not happen 

Gorg., 503 E. t^ ^^.j^g jj^^ ^^^^ j^^^^ together. The Laws, 773 A B, does. 

"*Cf'rphj iAA,Aa four times in 283, 284 with The,Btet., 482445 D. It cannot be a democracy, because *>Wo*o. 

160B,182B,Porme».,164C. _ _ ^^-^^^ iSvWror tlva. = Polit., 292 E, m»i' oiv iowl 

475 274, 275. *76 Gorg., 517 B, 521 T}, wA^Ods ye ev iroAet TavTriP ry]v iirtar^fxriv Swarhv elvai KTntraff&tu, 

«'304D, £««Aj/dem., 289DE. Cy. Gorp., 464-6, 502 E. 433 iJep., 338 D. Pindar., Pj/<A., 11, 87. 

478304E, Euthyde!m.,2Sa'B. iHiPolit., 291, 301, 302Cff. 

Paul Shoeey 63 

descending scale in the Republic. This leaves no place for a good form of democracy 
or a good monarchy apart from the ideal kingdom."* The fundamental distinction of 
the scientific state once noted, Plato plays freely with the conventional terminology, 
and no inferences can be drawn from his " contradictions." There are countless forms 
of government if one cares to look beyond the conspicuous eoSr].*^^ In the Republic 
the good oligarchy, the aristocracy of the Politicus, is a timarchy. In the Menexenus 
the good democracy of Athens is an aristocracy governed by kings!*" In the 
Laws,*^ from the historical point of view, all governments are regarded as variations 
of the two mother types, the Persian absolutism and the Athenian democracy. But 
in respect of the ease with which reform may be effected the tyranny ranks first, the 
kingdom second, a certain type of democracy third, and oligarchy last.*" I have 
already discussed the significance of the opposition of the two temperaments for the 
definition of the virtues and the antinomies of the minor dialogues.*'" Grote strangely 
ignores this when he affirms that these difficulties are not touched in the Politicus. 


The Philebus was selected by Dionysius of Halicarnassus as a type of Plato's 
simpler Socratic style. The majority of recent critics more plausibly see signs of 
Plato's later manner in the poverty of the dramatic setting, and the curious elabora- 
tion of phrasing and logical framework. The introduction presents again the objections 
to the theory of ideas advanced in the Parmenides, and, like the Parmenides, but 
more explicitly, hints that these puzzles are due to the limitations of human reason.*" 
It bids us disregard them and, assuming ideas, to deal with them and our subject 
according to the true dialectical method set forth in the Phcedrus.*^ It does not 
state that these metaphysical problems must be solved before we can so proceed.*'' It 
merely says that we must come to such an understanding about them as will prevent 
the puzzle of the one and many from confusing our inquiry.*'* We have no reason to 
look for a solution of them in the subsequent course of the argument. None is given. 
There was, as we have seen, none to offer. *°* The attempts of modern scholars to find 
one are very ingenious.*"' But they are not supported by Plato's words, and they 
proceed on the erroneous assumption that he thought it possible to give any other than 
a poetical and mythical account of the absolute, or to say more of the noumenon than 

485 Tho Politicus does not describe the development of 4M Supra, u. 70. *'3 15 c, 16 A B. 

one form from the other but merely states the order of tu (y_ qq this point my criticism of Jackson, A. J. P., 

preference among the lawful and lawless forms of the \gi^ jx, pp. 279, 280. Even Schneider (Plat. Metaphyeik, 

three types. Campbell, Intr., p.. xliv, overlooks all this p_ 53), whose interpretation of this part of the Philebus is 

when he treats as proofs of lateness the addition of excellent, does not make it clear that the metaphysical 

fiaiTi,\eia. as One of the lower forms, and the depression of problem is merely evaded by the assumption of ideas and 

oAiyapxiA below SrjfioKparia, the method Kar^ etfiij. 

<86 iJep., 544 D. 487 238 D. 488 693 D. 4955„pm, p. 36. 

489 710 E. The paradox, rvpovroujienji' /»oi {ore riiv iroAii', 496 As types of all may be cited : Schneidbe, Platonische 

709E, is literally incompatible with the associations of Metaphysik; Siebbck, UntersiKhungen zur Philosophieder 

Tvparvot in the Bepublic, but the notion of a revolution Oriechen, II; Plato's Lehre von der Materie; Henett 

accomplished by arbitrary power is found in 501 A, 540 E. Jackson, Plato's Later Theory of lieas. See A. J. P., 

490Stipro, pp. 11, 13, 15, n.59. 491 Stjpro, pp. 36, 37. Vol. IX, p. 282, 

64 The Unity op Plato's Thought 

that it exists.*" The elaborate apparatus of classifications and categories employed to 
decide whether pleasure or intelligence is more nearly akin to the good is due, apart 
from Plato's interest in dialectical exercise, to his unwillingness to treat the problem 
of the good in isolation. His imagination and religious feeling require him to associate 
the ethical good of man with the principles of order, harmony, measure, beauty, and 
good in the universe. We thus get many interesting analogies with the Timceus, but 
no solution of the problem of ideas. The direct classification and estimate of the 
different species of pleasure and intelligence, which was all the ethical problem 
required,*" is subordinated to a larger classification of all things which, however, 
deepens and enriches our conception of the psychological and ontological relations of 
the elements of merely human good and happiness."' 

The terms of this classification are the Tre/aa?, the aireipov, the /mktSv or mixture of 
the two, and the air^a or its cause. These terms represent, for the purposes of the 
argument, characteristic Platonic generalizations*" of the ideas naturally associated 
with these words. Whatever else they may mean is at the most suggestion and 
analogy. Ile/Jo? is a generalization of the idea of limit — whether it be the limitation 
of matter by form, of chaos by the principle of order and measure, of appetite by 
reason, or of the indetermiaate genus by a definite number of species and sub-species. 
It is the idea of the Timceus, so far as that is conceived as a priuciple of limit and 
form stamped upon chaos. But it is not the Platonic idea — the hypostatization of the 
concept — for the purposes of metaphysical theory."" 

The aireipov denotes among other things (1) the indefinite multiplicity of particu- 
lars as opposed to the unity of the idea — a conception found elsewhere in Plato.^ 
Plato generalizes the term aStfia for "matter" in 29 D. (2) Indeterminate matter as 
opposed to the form or limit that shapes it. In this sense it may be "equated" with 
the space, matter, or mother of all generation in the Timceus, 50 D.'°' (3) Indeterminate 

*^T Cf. Emeeson, Bepreeentative Men, "Plato," "No 603 Tftecetet., 147 D, tireiS)) aireipoi to irAiffos . . . .(vXXaPttr 

power of genius has ever yet had the smallest success in «is fv implies the method of Phileb., 15, 16. Cf. Sep., 525 A; 

explaining existence. The perfect enigma remains. But Polit, 262 D; Soph., 256 E; Parmen., 158 C. SoHNEEDEE, 

there is an injustice in assuming this ambition for Plato." p. 4, n. 1, notes this meaning, but still insists that the 

«8The net result of the introduction is (19 B) eiSij a7reipo» of the Philebus primarily means indeterminate 

■yap (lot SoKi'i vvv i/uaT^v riSovrit rfiiK 2uKpan)5, etc. matter, which he rightly shows is not = it-v ov, p. 5 (cf. 

*89 23C ff tupra, n. 261), but wrongly denies to be virtually identical 

600 So Plato generaUzes ixix,, Euthyd., 271, 272 : k^A,o« ''"^. ^P^"!- See Siebeck, p. 84. The Timmus does not 
(inM,' «v.T,), ibid., 289 E ; Awcvtk^, ibid., 290 B j Laws, "Pli^Uy identify matter " and " space merely because 
823B; Polit, 2990; Bep., 3733; Soph, 221, 222; .rA.o.efia, it does not distinctly separate the two ideas. SeeA.J.P 
Laws, 906 C, cf. Symp., 186 C, Gorg., 508 A; «tp6,W, Vol. IX, p. 416. But whether we call it matter or space, the 
Laws,mD; noir,„.,, Symp., 205B; lp<05, idid., 205 D, and X"P«, the -ra.{.x«, the mother of generationis one. 
passim; yiviaK, Polit., 261 B, etc. ; JiaitpiTiK^, Soph., 226 C ; 603 SiEBECK compares it as the antithesis of the idea to 
irtSai'ovpYiicq, ibid., 222 C; KoKoKtia, Gorg., 463 BS.; the the |;i>i oi', the crepoi' of the Sopftiat, the matter or space of 
comparative degree, rh ii.a\K6v re ital !ittov, Phileb,, 24 ; and the Timcetis, the principle of necessity or evil, and the 
many minor examples, Polit., 279, 280, 289. ne'ya /cni (xiicpdi/. More precisely (p. 89), the oVeipoi' is the 

601 ScHNELDEE, p. 133, and Siebeck, p. 73, make it a mediating link between the Oirepov of the Sophist and the 
mediating principle between the idea and phenomena. x^P" of the Timceus. Now these terms undoubtedly have 
But Platonever speaks of the "idea," but only of the ideas this in common, that they are variously opposed to the 
or the idea of something, nipas is itself an idea and is the ideas, but Plato employs them in different connections and 
cause of limit, in any given case, precisely as the idea of we cannot equate them. Siebeck argues (pp. 58 ff.) that 
whiteness is the cause of white, or the idea of dog the the absolute (i>i ov abandoned in the Sophist (258 E) must 
cause of a dog. mean something. He finds it in the absolute hypothesis of 

Paul Shoeet 65 

physical and chemical "process," as opposed to ideally or mathematically defined 
"states." ^ (4) The insatiate, limitless character of undisciplined desire and appetite — 
a conception which we have met in the Gorgias.^ 

The fUKTov is the mixture or union of 7repa<! and aireipov in any or all of these 
senses giving rise to various jeveaeK, both in the world of matter and in souls. '^°° As 
the union of matter and form it may be "equated" with the "offspring" of the idea 
and the "mother" in the Timceus.^^ As the mixed life of pleasure and intelligence it 
obviously may not."*' 

Alria is the principle of cause in general, and in particular the cause of the due 
mixture of pleasure and intelligence in the happy Itfe.^ In the one sense it may be 
identified with the Demiurgus who embodies the principle of cause in the Timceus.^^" 
The ultimate cause is conceived by Plato as beneficial intelligence which is virtually 
synonymous with the good. He intentionally confounds the good in human life with 
the good in the universe. It is possible, then, to say that God, or the good, or 
beneficent intelligence is the cause alike of the cosmos or ordered world and of the 
well ordered life."* We may identify the supreme mind (i/oiJ?) with the Demiurgus of 
the Timceus and the Idea of Good in the Republic. We may conceive the ideas as 
thoughts of God, identify God with the sum of his thoughts (vorjaK vo^o-ews) and so 
bring the ideas under the principle of ahia as not only formal but efficient causes."^ 
But in all this we are mechanically "equating" the terminology and imagery — the 
literary machinery, so to speak, of three distinct lines of thought in three different 
dialogues, for the sake of attributing to Plato a rigid and ingenious metaphysical 
system wholly foreign to his spirit. 

We have already discussed the psychology and the main ethical argument of the 

the Parmenidee as the antithesis of -the ev regarded as the ment and measure are spoken of in connection with the 

symbol of the principle of the ideas. From this it is an ideas, and movement and measure imply space ! 

easy step to identifying it with matter which is also the soiPWJefc., 24 B, 25 C, 26 A. 

antithesis of the idea. But it is not true that the absolute 605 27 E, 31 A, Gorg., 492-4, supra, p. 24. 

^r, iv must mean something. Plato's rejection of it in the 606 27 D, 25 E, 26 B, «al iv ifr«xa« aJ ni^wokka, which alone 

Sophist is sincere, and is confirmed by the Parm^tdes refutes the equation, irreipoy = matter. 

which makes it unspeakable and unthinkable. The abso- 60750D 

lute oi-, as we have seen, was reinstated for religious and Eoonii, ' • i- v.. • i- ..i. j.- 

. t • I -i. ■ \. vi u - «« ocs There is a slight equivocation in the assumption 

metaphysical purposes, as it is by many philosophers of ,„„t\\ j.i, ^ ^v .. • j„i-« « 1 j • 1 n- 

*^ „, , I- t t ■ (27 D) that the ' mixed " life of pleasure and intelligence 

every age. There was no such motive for forcing a mean- i. , . ,, . « . j . 

ing upon the absolute m 5.", and the identification of it ^^"""^ *° ^^^ '"""■' °* "'!•'" ^""^ "'"'""'• 

with matter is, as we ha ve seen, quite impossible. (Supra, a>i'.,i^u,m\j. 

jj 261.) 610 In SOD the /StwriXi/tijK 'fnixn", etc., = the soul of the 

SrEBECK then proceeds to associate the logical i-nupav world, and the airias Sivay.iv = the Demiurgus. 

and the Bartpor with space and to attribute to Plato an "' Cf. Idea of Good, pp. 188, 189, n. 2. 

" intelligible " as well as a phenomenal space by pressing E12 Schkeideb identifies God not with the Idea of Good, 

all passages in which the logical relations of concepts are but with the ideas. The ideas, he argues, must be real and 

expressed in spatial terms (p. 90). As the human mind they must be thoughts. They are, therefore, thoughts of 

naturally thinks logical determinations in spatial im- God. We have already considered this theory, sMpra, p. 38. 

agery, he has no difficulty in finding such passages. But It is for the modem systematic philosopher the most 

plainly the method is vicious. We cannot infer an Intel- plausible escape from the difficulty of positing two dis- 

ligible " space " or the identity of Sartpov and space tinct noumena, God and the Ideas. Perhaps Plato would 

because the ideas are spoken of as '^living apart," or have accepted it, if it had been presented to him. Unlike 

" included" in a larger idea, or because the method of the majority of its advocates, Schneideb does not misin- 

dichotomy proceeds to the right and leaves on the left the terpret particular passages in order to support it. He 

other of the particular idea pursued. Still less c£in we merely combines and equates lines of thought which Plato 

infer it from the vtntrot ro-nat, or from the fact that move- left unfinished and distinct. 

66 The Unity of Plato's Thought 

Philebus, and seen that neither contradicts or appreciably modifies the doctrine of the 
earlier dialogues."' There remains only the question whether the demonstration of 
the unreality of pleasure presupposes, or, as Zeller still maintains, is presupposed by, 
the shorter proof of the Republic. Believing that the Philebus is probably late, I am 
logically committed to the first branch of the alternative. But this opinion is entirely 
compatible with the view that the difiPerences between the two treatments of the theme 
are not in themselves sufficient to show which must be the earlier. It is impossible 
to determine a priori whether the slighter treatment is an anticipation or a r6sum6 of 
the fuller discussion. The main doctrine was always a part of Plato's thought, as 
appears from the Gorgias, the Phcedo, and the Phcedrus."* The differences between 
the Bepublic and the Philebus have been much exaggerated. The abbreviation of 
the argument in the Republic is sufficiently explained by the subordinate place which 
it occupies in the scheme of the entire work. It affords no proof of the date, and no 
presumption even of a change of doctrine."^ 


The date of the Thecetetus has been much debated on external grounds.'" Its 
wealth of thought and dramatic vivacity of style make it one of the most difficult 
dialogues to classify. In psychological depth and dialectical acuteness it ranks with 
the Sophist, Philebus, and Parmenides, many of the thoughts of which it anticipates 
or suggests.'" But it has nothing of their dogmatic finality of manner. Socrates is 
still the midwife delivering ingenuous youth of opinions which fail to stand the test 
of the elenchus. And the conclusion is an avowal of Socratic ignorance. ''* 

Before losing ourselves in details we must recall why this is so. There are two 
reasons: (1) The formal quest for an absolute definition always fails in Plato."' (2) 
It is not possible to define knowledge or explain error. We can only describe and 
classify different stages of cognition and various forms of error. All seemingly intel- 
ligible explanation rests on material images, like Plato's figure of the wax tablets and 
the aviary. But these analogies either commit us to sheer materialism and the flowing 
philosophy, or they explain nothing. No spatial image can represent the synthetic 

513 Supra, pp. 24, 43, 45 ff. that the Bepublic is not yet acquainted with the thought 

^^^ Supra, p. 24. that the neutral state implies not absolute quiet in the 

■ina ^nr.nn Kjo rri, f I, iv, 1 body, but slight motions which do not cross the threshold 

»i» See Zellee, p. 548. The question whether pleasure , . Tii^vit ti-- i-j.„ ^^ 

. . • ... J ,„ cnj!T>\ A I. %. o or consciousness. But the thought is implied in ifec. Of. 

or <l>pofri<TK is the good (Sep., 505 B) need not be a specific a ^ i- j 

reference to the Philebus. It is virtually raised in the ^ i • 

Protagoras and Gorgias. Zeller's table of agreements be- *'° See Zelleb, p. 406, n. 1 ; Campbell's Introduction j 

tween the Bep. and Phileb. merely proves the unity of Lutoslawski, p. 385. It is on the whole more probable 

Plato's thought. B^., 584D-585A-E, 586A-C, which he that the battle in which Theaetetus was wounded belongs 

cites, present, at the most, different imagery. The thoughts to the Corinthian war, 394-387, than to the year 368. 

are in the Philebu^s, That the Philebus does not refer spe- 517 Qf supra pp. 33 34 55 nn. 179 182 389. 

cifically to the Idea of Good is no stranger than is the fact 

that no other dialogue does. On the other hand Ldtos- ^"^^^ ^■' 161 AB, 209 E, 210 C. 

LAWSKi's objection (p. 470) that the difficulty, iJep., 505 B, 619 0/. supra, p. 13, p. 16, n. 86. JowBTT says. Vol. V, 

that the sought-for (^pdi'ijo-t? is </ipdi'7j(rts toO aya.6ov is disposed p. 119 : " We cannot suppose that Plato thought a defini- 

of by our observation that the reference, if reference there tion of knowledge to be impossible." But it is impossible, 

must be, is to the Charmides, supra, n. 61. Jackson argues and that for the very reasons suggested by Plato. 

Paul Shobey 67 

unity of conBciouBness and memory. None can explain the comparison of past and 
present impressions in an unextended focal point of consciousness. None can repre- 
sent except in the vaguest poetic figure'^ a psychical mechanism that now operates 
correctly, yielding right opinion, and now incorrectly, resulting in error/" On the 
other hand, if we iavoke the absolute unity of mind behind our imagined mechanism, 
we are merely moving in a circle. We reaffirm our faith in the immaterial soul, but 
we can offer no intelligible explanation of degrees in cognition or of the psychological 
process of error.^^^ 

The quest for a definition, then, fails, as Plato expected it to do. But the analysis 
is carried far enough (1) to refute to Plato's satisfaction all psychologies of pure mate- 
rialism or relativism ;^^' (2) to justify a purely logical and practical treatment of the 
fiT} 6v, yjrevBr)'} So^a, and similar fallacies in the Sophist.™ This and the immense 
wealth of psychological suggestion scattered by the way are the chief positive results 
of the dialogue.''' 

It has been repeatedly analyzed in detail.'^" As in the Gorgias and Fhilebus,^^^ 
much of the argument is purely dramatic, directed only against the cruder forms of 
the theory combated.'^' The ingenious attempts to reconstruct the doctrines of con- 
temporary thinkers from Plato's polemic are more apt to confuse our understanding of 
Plato than to add to our knowledge of Protagoras, Aristippus, or Antisthenes.''" As 
Professor Campbell says: "Whoever the contemporaries were to whom Plato refers as 
the disciples of Protagoras, he aims beyond them at the whole relative side of Greek 
thought of which Heraclitus was the most prominent exponent." 

The identification of the dv6pcoiro<! fierpov, the trdvra pet, and the definition that 
knowledge is sensible perception, is a part of Plato's literary machinery which we 
must accept untroubled by nice historic scruples. Th^ dvOpmiron fierpov is not a 
scientific or philosophic principle, but a rhetorical paradox or truism embodying a 

S2U c/. Tim., 37 AB, with Thecetet., 194 B. 624 Cf. supra, p. 55. 

621 Zellek, p. 590, thints that the section on i)<6uS)i! Sofa 525 On its relation to the theory of ideas cf. supra, p. 33. 

is an indirect Testation of the definition that knowledge ^^ ^^ Campbell, Joweii. Gkote, etc. 

is aA7)9iis «o|a. He says that the diflSculty of eiplaining 

false opinion arises only from the assumption that knowl- 527 Supra, u. 137. 

edge is "right opinion." That is not so, either absolutely ,, , , 

. ", . „,, ,i. i JO. li. • 'jii. -J — ,„«„ 628 Supra, n. 7. Note especially the tone of 16S-6, 

or in Plato. The ultimate difiBculty is : if the mmd appre- j, • i- .. 1 j ■ 1 

or lu jTiauu. o-uo uj„.ii.o„ j „,^;„ where aTowedly eristic arguments are employed against 

Aerute as a psychic unit, how is m».-appreAer««m as dis- ^^^ ^^^^^^ identification of i...r^M and a:.«,a«. Observe 

tinguished from non-apprehension, possible 1 BONITZ .s persiflage of 156, 157, 167 A, 179, 180. Natoep, Philol., Vol. 

undoubtedly right in affirming that the question for Plato l, p 263. thfnks 161 B-iek E a Urody of Antistheues's attack 

is not so much the fact or POSSibJity of «"°-^=the psycho- ^^^^ ^^, C being Protagoras's defense. Any allu- 

logica explanation. (Pp. 83, 89 Cf. my paper DePi^ ^.^^ ^ ^^.^^^ ^^ .^ ^ ^^^^^ ^ p^^^^ ^^ Antisthenes 

torn. »<Jear«m doc*rt™. pp. 17-19.) The length of the „,„, j^er eristic contemporary. Protagoras himself is 

"digression" is justified by the interest attached to the ' 1 • it - - .-ki-i. iota rw 

» " J J V 1 • I 1 »• *!,«(. represented as employing the (i>) ox quibble, 167 A. Qf. 

problem of ifrevSiit Sofa and the psychological analysis that "v^-''^ j e, »i j oocri 

prooioiiiuir " .,,. -I, .1 „ „.„„n„„ ,.c„it supra, n. 405, and j;«{fti/dem., 286 C. 

it provokes. It is a "digression" and a negative result j- 1 • 

only for those who naively assume that Plato himself ei- 629 See Natoep's acute Forschungen zur Geschichte des 

pected to reach a positive definition. Erkenntnissproblems im AUerthum, and his "Protagoras 

622184CD 200AB. und sein Doppelgftnger," Philologus, Vol. L, pp. 262 S. 

623Sitpra,'p.34,n.283. Cf. Theaitet.,\HG S. UptolSSC Natoep's analyses. retain their value, even if we doubt 

the identity of iiT^iT-i,y.-r, and atcrSijo-it is refuted only so far the possibUity of reconstructing»Protagoras. For Antis- 

as it depends on extreme Protagorean relativity or Hera- thenes and the TheiBtetus see the phantastic conjectures of 

cliteanism, which makes all thought and speech impossible. Joel, Der echte urui der xenophontische Sokratet, Vol. II, 

Kara y€ TJjy tov Ttavra Ktxclo-Oat ttido^ov. PP- °^ "• 

68 The Unity of Plato's Thought 

practical tendency of the age repugnant to Plato's taste and feeling; This seems to 
be overlooked in the controversy between Natorp (Philologus, 50) and Gomperz, as to 
the meaning of the formula. Plato, as Natorp shows, explicitly affirms the thought to 
be: things are to (each and every) man as they appear to him. If sugar tastes bitter 
to the sick man, it is bitter to him — there is no other test. But there is no evidence 
and no probability that Protagoras had systematically drawn out the consequences of 
generalizing this proposition in its application to ethical and logical truths. He did 
not need to ask himself whether he meant by dvdpanro^ this, that and the other 
man, or human cognitive faculties in general. He took ovra, as he found it in Greek 
idiom, without distinguishing things, qualities, and truths — though his simplest 
examples would naturally be qualities. By o>? he presumably meant "that," but 
"that" and "how" are closely associated in Greek idiom and are often confounded in 
popular not to say in Platonic usage. If he used ^aiverai and {ftavraa-ia he probably did 
not distinguish the "it seems to me" of actual. sensation from the "it seems to me" of 
any opinion,^*' and Plato avails himself of the ambiguity for the half serious •jrepiTpoiri] 
that since Protagoras's "truth" does not seem true to the majority, it is admitted by 
Protagoras himself to be oftener false than true.'^' 

ndvra pel Plato himself accepts for the phenomenal world."' As a metaphysical 
dogma it is tantamount to materialism in that all materialists are more or less con- 
sciously Heracliteans, though all Heracliteans need not be materialists.''' As a neo- 
Heraclitean paradox it is the negation of the ideas, of the universal, of rational logic 
and speech.^'* As a rhetorical formula it is the symbol of the restless spirit of innova- 
tion which Plato detested.''" Before generalizing and restating for serious refutation 
what he conceives to be the common psychological presuppositions of these catchwords, 
Plato covers them with persiflage and assails them with arguments which he admits to 
be rhetorical and eristic. There is no probability that the representatives of these 
doctrines could have explained their meaning or defended themselves as well as Plato 
has done it for them. So far as we know, he is the first thinker who was capable of 
distinguishing, dividing, classifying, and generalizing ideas, of noting the affinities 
and differences of philosophic doctrines, and of translating them freely into different 
terminologies. All other early thinkers, like the majority of thinkers always, are the 
prisoners of their formulas and can only aboimd in their own sense. Plato, as 
Emerson says, "needs no barbaris war paint, for he can define and divide," and he 
delights to prick with the keen point of his dialectic the bubbles of imagery, rhetoric, 
and antithesis blown by his predecessors. Heraclitus means well when he says that 
the one is united by disimion,"° or that the hands at once draw and repel the bow."' 
But the epigram vanishes under logical analysis. The pre-Socratics discourse, in a 

630 Cf. supra, p. 48. 63* Cratyl., 439, 440 ; Thecetet, 179, 180 ; Soph., 249 D. 

531 no, 171. Cf. Euthyd., 288 C, taX tov's re a AAou! avinpinoiv 635 Pateb, Plato and Platonism, pp. 16-20. 

632Cra<2/Z., 439 D; Si/mp., 207 D; KTmiBiw, possim. „, „ .„„ _, . . „ ,. 

637iJep 439B. The saying is Heraclitean in tone. 
533r7iecEte«., 155E,156A. 

Paul Shobey 69 

fine imaged style, about Being, but a plain man can not be sure of their meaning. ''* 
Absolute formulas, like irdvra pel, irav ev, Travrmv fierpov avBpcoTro^, have an imposing 
sound, but if we press for their interpretation, prove to be either truisms or paradoxes, 
destructive of intelligible speech. °^' 

It is an ingenious sport to construct for Protagoras some subtle and nicely 
guarded modem system of phenomenalism. But we must then pass over the purely 
dramatic parts of Plato's discussion, and limit ourselves to his final and seriously meant 
arguments against the psychology of materialism and the logic of relativism. There 
are two such arguments which neither Plato nor his critics are careful to distinguish 
sharply: (1) The first is that the senses are organs of mind and that sense perception 
itself implies the "soul" or some central "synthetic unity.""" This, if fully under- 
stood, is conclusive against the sensationist materialism of Condillac's statue. But 
Plato's chief interest is in the second argument derived from this. (2) The objects 
of each sense we can perceive only through the specific organ of that sense.^' But 
the general common categories of Being, not-Being, number, likeness, difference, the 
same, and the other,"' as also ethical universals, and the abstr-act definitions of sensuous 
qualities °" are apprehended without subsidiary organs solely through the action of the 
mind, and its reflections on the contradictions of sense. Availing himself of the 
double meaning of ovaia (1) logical essence, (2) reality, truth, Plato argues, as in the 
Phcedo,''^ that truth and reality are attained only by the "pure" thought of the soul 
acting independently of the body. 

A modem Theaetetus, of course, might deny that abstract thought has no bodily 
organ, or that its objects are more "real" than the perceptions of sense. But the 
absolute identification of ato-^jjo-t? and ewiarrifiri is sufficiently refuted, and the suggest- 
iveness of this definition having been exhausted, a fresh start is made with the 
definition "knowledge is true opinion." But this implies that we understand erroneous 
opinion, and error proves to be inexplicable. The attempt to explain it calls forth 
many interesting analogies and distinctions. '** One large class of errors is accounted 
for as arising from the wrong reference of present sensations to stored up memory 
images.'^ The distinction between latent or potential and actual knowledge postpones 
the final difficulty."' But in the end it must be faced: error as a matter of fact occurs 
in "pure" thought. How can pure thought misapprehend its object? A bodiless 
intelligence either touches or does not touch the object of thought. We can understand 

638 Soph., 242, 243. nerves, but Empedocles " already " remarked of the senses, 

6390ra{jl., 439,440; Thewtet., 183AB, n9DE; Soph., oi Svvatreai n HA^Kav Kpivciy, Theophr. sens.,!, Dox. 500. 

249CD. 542185CD. 6« 186 A B. C/. supra, nn. 221 and 222. 

6*0 184 D, «"r6v yip ^ov, S, ir«i, « iroWiii tc«5 iv ji^lv, S,a^,p ^^^ y^^^j^j^ jg, ^ . pj^,^^^ gS C. 
iv SoupEiow t;riro«, atffS^ireit eyicaftiji'Tot, aXAa (tij «i! /iiov riva 

iSiav, .Ire ^xX" "Te 5 « &,l Kol^lv, nivra TaSra {v-ret-et, etc. '" Of- '"i*^"' V.55;u. 520 with text. 

6"185AC. LUTOSLAWSKI, pp. 276, 372, fancies that this . 6««193,194. The memory image is treated as knowledge, 



is an anticipation of the modern " law of specific energies 
of the senses," " already " glanced at in Bep., S52E, but 
showing progress in the formulation here. The modern 
law could not be anticipated without knowledge of the »iii eiSiviu. 

of the senses," " already " glanced at in Sep., S52E, but 5*7197. This is the distinction invoked in JJutAyd., 277, 

showing progress in the formulation here. The modern 278, to meet the eristic fallacy of the alternative liJfVat % 

70 The Unity op Plato's Thought 

the confusion of one object with another, the misplacement of cognitions, only in 
terms of spatial imagery which, if accepted literally, is materialism again, and if taken 
as a symbol implies the synthetic unity of mind behind it, and so renews the puzzle in 
infinite regress."* Modern metaphysicians evade the difficulty by assuming aa infinite 
thought of which our erring thought is a part. Their task then is to preserve the 
iudividuality of a consciousness that is part of another mind. This problem disappears 
in a mist of theistic language enveloping pantheistic doctrine. Plato does not soar to 
these heights, but having carried the psychological analysis to the limit, he disposes 
of the equation, eTna-Trjtirj =^ \6yo<i akrjdiji, by pointing out a sharp practical distinction 
between knowledge and right opinion. True opinions may be imparted by persuasion 
and hearsay about things which we can^know only if we have seen them."' 

The third and final suggestion is that knowledge is right opinion coupled with 
Xo'yo?.*™ This is for practical purposes substantially Plato's own view."' Tran- 
scendentally knowledge is the apprehension of the idea. In human life it is the 
dialectician's reasoned mastery of his opinions implying stability, consistency, and the 
power to render exact account of beliefs. Plato reserves the terms knowledge, 
intelligence, pure reason, for the man who co-ordinates his opinions, unifies them by 
systematic reference to higher principles, ideals, and "ideas," and who can defend 
them in fair argument against all comers.**^ This is not a definition, but it is quite as 
good a description as the most modern of his critics can produce. This view is set 
forth in the Republic in the context necessary to make it intelligible. It would not 
have suited Plato's design to repeat or anticipate that description in the Thecetetus 
which is cast in the form of a dialogue of search. Moreover, it is one thing to give a 
general definition of knowledge and another thing to describe the state of mind to 
which the term science or knowledge kut i^o^riv is applicable. Sensible perception 
is not a synonym or definition of knowledge, nor, according to Plato, knowledge in the 
highest sense. But it is the most certain and the only knowledge we possess of some 
kinds of objects. And the recognition of this fact in various passages of the Thecetetus 
would in itself make a satisfactory all-inclusive definition of knowledge impossible."' 

Accordingly Plato brings the dialogue to a plausible conclusion by discussing 
(and rejecting) various possible meanings of X6yo<;, none of which yields a good defini- 

S48 20OAB. The original ajropCa arose from the nnme- exeii- Aoyoi' SoCi-ai is opposed to eirurT^fiT). In ethics fixed, 

diated antithesis eifieVai ^ firi cifieVai — a conscious fallacy, stable, true opinion is virtually a synonym of ^poi'ijtrts : 

as the language of 188 A and Euthyd., 277, 278, shows. Psy- Laws, 653 A, ^/lovTiaiv it «ol aAi)9eii Sofas /Se/Saioui. Strictly 

chology is enriched, and the practical fallacy is disposed speaking, there are three grades : (1) casual right opinion ; 

of, by the distinction of grades and kinds of cognition, but in (2) right opinion fixed by judicious education from youth ; 

the end our analysis brings us to an indivisible act of (3) right opinion fixed and confirmed by the higher educa- 

psychic apprehension which either is or is not. tion and accompanied by the ability Sovvai. X6yov. But Plato 

549201B; Grote triumphs in the admission that sense- is not careful to distinguish the last two. They are both 

perception is, after all, sometimes knowledge; cf. supra, li-ovifLoi (Mew), 98A; iJep., 430B, reading/idi'i^oi'). InPoJif., 

n. 324. 309 C, a.K-q9r) 56^av /x€Ta )3e)3ai(ucre(i)s cannot be referred exclu- 

650 201 CD. sively to the philosophic virtue with Zellee (p. 5%). It 

551 The Timceus (51 D) sharply distinguishes vom and includes the virtues of fixed habit guided from above, as 

iA7|9i|! Sofa, but adds TO ii.iv i.A /leT aATiSou! A070U, TO ie a\oyoi>. appears, e. p., from the reservation ill ve ev iroAireio, 309 E, 

In the Meno, 98A, right opinions became knowledge when which is precisely equivalent to ttoAituc^i' -/<: in Sep., 4S0E. 

bound oirias Aoywrj^y. InSymp., 202A, opOaSofa^eti- — avcvroO 5525itpya, p. 17; n. 91 with text. BSSfif-jipj-a, n. 549. 

Paul Shoret 71 

tion/" Socrates has heard a theory that the first elements of things are simple and 
not objects of knowledge. — Eqt know ledge imp lies giving and taking an account, and 
no account can be given of elements beyond naming them. They^wHTnot admit any 
other predicate."^ In this paragraph we may discover allusions to Antisthenes's paradox 
about predication and definition, to current philosophies of materialism, and to 
mechanical interpretations of Plato's own formula Sovvai re koX Be^aaOai Xoyov. But 
whatever Plato's secondary literary intentions, his main purpose is to present a 
serious psychological and metaphysical problem. Is the whole the sum of its parts 
except in mathematics? Can the world be explained as a mechanical summation of 
elements? The problem presents itself to us in psychology and cosmogony. *'° Plato 
treats it in dialectical abstraction, taking the syllable and its letters ("elements," 
a-Toi)(^e2a) as representatives of elements and compounds. He decides (1) that the 
syllable is not the mere equivalent of its elements, but a new emergent form and dis- 
tinct idea ; (2) that, whether this be so or not, the elements and the syllable are equally 
knowable and unknowable. For if the syllable is the sum of the elements it cannot 
be known if they are not. And if it is a new unity it is as elemental as they and can- 
not be explained by resolution into its parts. 

The second conclusion disposes of the proposed definition. The first, as we have 
already seen, is a suggestion of the doctrine of ideas as against philosophies of 
mechanical materialism.*" But we are not therefore justified in making this episode 
the chief purpose of the dialogue. Two other possible meanings of XSyo's are shown to 
yield no result, and the dialogue closes with the Socratic moral that we are at least 
wiser for knowing that we do not know. 


The Phcedrus, with its profusion of ideas, its rich technical and poetical vocabu- 
lary, and its singular coincidences with the Laws"'^ and Timceus,^^^ makes the impres- 
sion of a mature work. This impression is confirmed by Sprach-Statistik, and by the 
fact that it directly parodies a sentence of Isocrates's Panegyricus published in SSO."*" 
It is possible to say that the thoughts are merely sketched in a "program" of future 
work ; that the dithyrambic vocabulary is due to the theme ; and that the phrase of 
Isocrates is taken from an older, common source.^' Anything may be said in debate. 

B5*.LtiT03LAW8Ki (p. 371) argues that the Thecetetue re- 5" 202. 

jecting Ao-yot, etc., contradicts the opinion "provisionally" ^^^E.g., Wcndt's psychology differs from that of the 

received in JfejM), 98A, Sj/mp., 202A, and PA(E<Jo,96B. He pure associationists chiefly in that he insists that the whole 

fails to note (1) that this "provisional" view recurs in the jj not the sum of its parts — iAV ef Utiymv ip ti Ytyoi/af t'Sot 

TimcBUS, (2) that Fhczdo, 96.B, is an ironical summary of ^ay liiav ovto airoS lx<"'i Thecetetus, 203 B. 
materialism and is irrelevant here, (3) that the omission of HT Supra, nn. 227, 228 with text, 

airia which surprises him (p. 378) is presumably intentional ^^ ^^ jj ^^^-^ wr^aeut, etc. 

and minimizes the contradiction. Plato does not intend to ^^^ ^^ ^^^ ^^ ^^ ^ 

"define" knowledge, but he is careful not to contradict the „„„„„., , - -, ., 

practical description of it given in the Bepubhc. The , "'\'^' , ^ \ J o ■ - -/ 

phrase Sovvai. re <tal Sefa<r9at \6yav is mentioned as a conditio *°'^'' ^' »PX«'^5. «>-«- ■<»•«-• ■^""- °. /^ t 

sijie guo rwwi of knowledge (202 C), but only m connection - , .- - ,. 

with the rejected theory of elements, and its fuU dialecti- "'"'"" «"'^»"''- «»<:• 
cal significance is not developed. '" Gompebz, Ueber neuere Plato-Forschung. 

72 The Unity of Plato's Thought 

But there is an end to all use of Isocratean parallels if we cannot infer that the Phcedrus 
is later than a work which it explicitly parodies. 

If we assume Lysias, who died in 378, to be still living, the date may be still 
more precisely determined to about the year 379. The strongest confirmation of this 
date is the weakness of the arguments for an earlier date, which it is hard to take 
seriously. The politician who recently called Lysias a Xoyoypd(j)0'i need not have been 
Archinos, and, if he was, Plato's use of evayyp'i may be merely dramatic.^*" The 
patronizing commendation of Socrates at the end°°^ is not incompatible with a sly 
parody of his Gorgian style, nor even with the sharp rap on the knuckles administered 
to him (if it is Isocrates) at the close of the Euthydemus. Still less can we say that 
Plato and Isocrates could never have been friends after the declaration at the close of 
the tract against the Sophists that virtue cannot be taught, or, for that matter, after 
any other polemical innuendo in their works. Huxley, Matthew Arnold, Frederick 
Harrison, Herbert Spencer, and other knights of nineteenth -century polemics, com- 
bined much sharper thrusts than these with the interchange of courteous or slightly 
ironical compliments. 

Our chief concern, however, is with arguments drawn from the thought. We 
have already seen that the dialectical method of the Phcedrus is not appreciably less 
mature than that of the Philebus or the Sophist,^^ and that, on the other hand, there 
is nothing in the psychology or ethics of the Phcedrus that necessarily fixes its rela- 
tion to the Republic, the Phcedo, or the Symposium.^ What can be said, then, of 
the attempts of distinguished scholars to show that the thought of the Phcedrus dates it 
circa 392, or even ten years earlier ? The only one that calls for serious consideration 
is Natorp's argument™ that the immaturity of the Phcedrus is proved by the absence 
of the notion of a supreme science, or of ultimate categories found in the Symposium, 
Republic, Sophist, and even in the Euthydemus. The answer is that such a notion 
never appears in Plato except in some special form adapted to a particular argument. 
Natorp includes very different things under this rubric. The supreme science of the 
Symposium is merely the knowledge of the idea — of the idea of beauty as distin- 
guished from particular beauties. That of the Republic is knowledge of the idea — 
of the idea of good as the o-kotto'; or aim of true statesmanship. That of the Euthyde- 
mus is in one place by implication dialectic (290 C), in another the "political art" 
(291 C). In other passages the unity of science is merely the unity of the concept or 
idea, eVtcrT^/iT;.^" The ontological categories of the Thecetetus, Sophist, and Parmeni- 
des belong to a different line of thought and have a mainly logical significance. They 
are connected with the notion of a universal science only in so far as they are appre- 
hended and discriminated by dialectic. Now the subject of the Phcedrus did not call 
for the explicit assumption either of supreme categories or a universal science. The 
chief point in the myth, ignored by Natorp and the majority of commentators, is that 

662 257 C . 564 Supra, u. 377. 565 Supra, pp. 19, 43 ; n. 152. 

663 279 A, Tout Adyows oil fiy <irix«ipei may well be the 566 Heemes, Vol. XXXV, pp. 405 ff. 
Panegyricus, but might be anything. 567 Supra, n. 460. 

Paul Shoeet 73 

the ecstasy of love is due to a speciality of the idea of beauty. Unlike other ideas, it 
is represented in this world by a not wholly inadequate copy, the sight of which recalls 
the beatific vision of the original.^ The proof of immortality requires only the 
categories of the self-moved and that moved by another.^" The absence of other 
abstract logical categories proves no more here than it does in the Laws. The method 
of dialectic is described in its relation to rhetoric, which is regarded as an art of 
deceptive dialectic or almost eristic."" There is no occasion for going back to ultimate 
categories or hypothesis beyond hypothesis. The subject about which it is desired to 
effect persuasion is the starting-point.™ The rhetorician's art is to bring this under 
a definition or category from which there is a plausible transition to praise or blame.™ 
So even in the Philebus the account of the true dialectical method starts from the 
concrete a-jreipov to be investigated, or the idea, the ev, that it reveals to inspection, and 
says nothing there of ontological categories, ultimate hypothesis, or a supreme science.™ 
The Philebus is not for that reason less mature than the Phcedo."* Plato cannot 
always delay to tabulate ultimate categories or to reaffirm the unity of science, whether 
it be (1) as dialectic, (2) as the vision of the idea, or (3) as the "political art." 

Natorp's other arguments merely confirm our main position by illustrating once 
more, and typically, the desperate straits to which an acute scholar is reduced in the 
attempt to date the dialogues by their thought. For example, there is obviously no 
connection between the remark that those who affirm that <^p6vq(Ti<; is the chief good 
are unable to define what <j>p6vrjaii (Rep., 505 B), and the enthusiastic declaration that 
if wisdom (^poVijo-t?) could be seen by mortal eyes (as beauty in some measure can) it 
would enkindle Seivov<i .... eptoTat (Phcedr., 250 D). Yet Natorp regards the first 
passage as a distinct criticism of and advance upon the latter. But the Phcsdrus pas- 
sage merely says that (j^pourjo-K, if we could only see it, would be still more lovable 
than beauty. It does not affirm it to be the chief of goods, and, if it did, need not 
for that reason precede the Republic, unless we are to say the same of Laws, 631 0.°" 

Again, in 245 C the unctuous phrase SetvoU fiev a'iri(no<;, ao^ol<i he -Tria-T^ is said to 
mark Plato's early, unscientific mood, because mature Platonism ranks knowledge 
above ttio-tk. But plainly a religious thinker may affirm the superiority of knowledge 
to belief and yet indulge himself in the ironical declaration that the "clever" will 
disbelieve, but the wise believe, his proof of immortality. Similarly in 247 C the 
statement that no poet has ever worthily sung the region above the heavens is taken 
to prove that the passage is Plato's first exposition of the theory of ideas. But such 

568 250 BCD. 669 2450. in common with the five categories of the Sophist, the 

, . , supreme science of the Symposium, or the viraleaw of the 

570 261 D with Sophist, 259 D. Rhetoric is generalized to _p^^^^ ^^^ ^^^ Republic. 
include dialectic and eristic, just as in Sopftist, 222, 223, !"»«- 4.!„„j t„ „i„;„i„,>,.t„«,sr. 

,, , * u i - i.1. u- -i; * *-k« 575 pTtoedr., 250 D, seems destined to misinterpretation, 

voupyiCT embraces all forms of rhetoric, the higgling of the ^ «u,u., ., wv i|> =«. „n.„„r*i.o, ..,,1 

„„IJ *,,„ T.„„i..,-. „t „f fl,« r>»r»=itB. and t,hB whole LCTOSLAWSKI, p. 339, misses the meaning altogether, and 

market, the Lucianic art of the parasite, and the whole 

HOEN, pp. 212, 213, actually takes Seii-ovs Ipwraj (understand- 

teaching and eristic of the Sophists. .^^ ^^'JJ^-^ ^ ^ ^^^ ^^^^ ^^ p^^^^,^ _.^^^^^ ^^^ ^^ ^^^^ ^^ 

571 263 D E. 672 265, 266 A. 673 16 C D B. ^^jj images of other ideas than beauty, and objects that 

574 The division of all things into ire'pas, airnpoi', ^i.kt6v, the passionate love of justice would be a good, since it 

and aWia is given in a different connection, and has nothing would not be exposed to sensual excess I 

74 The Unity op Plato's Thought 

a prelude is a mere commonplace of rhetoric, as in Phcedo, 108 C ; Meno, 239 C ; 
Polit, 269 C. 

The argument that dialectic is first introduced as a new term in 266 will not 
bear scrutiny. In Philebus, 53 E, eveKci tov is introduced still more circumstantially. 
The ideas are a dream in Cratyl., 439 C ; dialectic is dramatically led up to in Cratyl., 
390; and in Sophisi, 265, 266, an elaborate explanation has to be given of what is taken 
for granted in the phrase (pavrdcrfiaTa Oeia, Rep., 532C."° Natorp says "der Begriff 
Dialektik ist im Gorgias noch nicht gepragt, sondern erst im Phcedrus." But 
StaXeyeaOai is contrasted with prjTopiKrj in the Gorgias, 448 D, and the term StaXe/crt- 
ic6<;-i], if I may trust my memory and Ast, does not happen to occur in the Symposium, 
Thecetetus, TimcBus, Parmenides, Phcedo, Philebus, or Laws. It is begging the 
question, then, to assume that Bia\ejea-6at in the Gorgias does not connote true Pla- 
tonic BcaXeKTiK'^, but only Socratic conversation. There is not a word about "damo- 
nischen BiaXcKTO'i" in Symp., 202 E, 203 A, and the notion of philosophy as the seeking 
rather than the attainment of knowledge occurs not only in Symp., 203 D- 204 B, 
"after" the Phcedrus, but in Lysis, 218 A. As for Xoywv rexvr), it is any "art of 
words," whether actual or ideal rhetoric, dialectic, or even eristic.^" It is uncritical to 
press the various meanings which different contexts lend to such a general expression. 
Rhetoric is called the Xdyav rexvrj in 260 D, but Socrates immediately adds that there 
is no true Xeyeiv Te^yr) avev tov aX'qOeCa^ rj<f>6ai; i. e., without dialectic. There is, then, 
no inconsistency between this and the use of rrj? irepl toxk \6yov<; Te^^wys in Phcedo, 
90 B ; nor can it be said that the \6y(ov iJ,edoBo<: of Sophist, 227 A, differs appreciably 
from the fiedoSo:; of Phcedr., 270 D."' Lastly, Natorp's argument (pp. 408-10) that 
the method of avvaywyrj and Biaipea-K described in the Phcedrus does not go far 
beyond the suggestions of the Gorgias and Meno is, of course, merely a further con- 
firmation of our main thesis. But when he adds that IBea is used vaguely in 237 D, 
288 A, 246 A, 253 B, etc., and not, as in the " later" iJepM&Kc and Phcedo, in the 
strict sense of Platonic idea, the reply must be that this vague, untechnical use of 
elSo? and ISea is always possible in Plato."" Omitting Thecetetus, 184 D, since Natorp 
thinks that also "early," we find it in Rep., 507 E; Philebus, 64 E, and Cratylus, 
418 E, where ayaOov IBe'a does not mean "idea of good." Since the transcendental idea 
is established for the Phcedrus, of what possible significance is the occasional use of 
the word IBea in a less technical sense ? 

These illustrations might be multiplied indefinitely. They do not establish a 
universal negative, but they certainly create a presumption against all arguments of 
the type which careful scrutiny always shows to be fallacious. And the experience of 
the untrustworthiness of many such arguments creates in the minds of sober philolo- 
gians a more justifiable "misology" than that which Plato deprecates in the Phcedo. 

5™ See Adam, ad loc. "s cf. supra, u. 377. 

'I'T Euthyd., 288 A, iiMtTdpat rexriJI .... oiT(aal 9aviiim)t H'SSee JoWETT AND Cahpbell, Vol. II, pp. 294 fl. 

outTTjs ets aKpipeiav Adywc. 

Paul Shoeet 75 


In vivacity and comic verve the Cratylus is "early," ^"' in maturity and subtlety 
of thought "late." Its most obvious feature, the playful allegorical use of etymol- 
ogizing, is anticipated or recalled in many other dialogues/*' Admirable is the art 
with which etymologies recognized to be little better than puns are made the vehicle 
of a true philosophy of language, and a profound discussion of the relations of lan- 
guage and thought. 

With this we are not concerned. We have already seen that the attempt to 
assign the dialogue an early place in the development of Plato's own thought breaks 
down.^ Plato is "already" in full possession of the theory of ideas and of the essen- 
tial arguments of his polemic against the flowing philosophers.™' His repudiation of 
eristic fallacies is as distinct and as clearly, if not as fully, expressed as it is in the 
Euthydemus and Sophist.^"* 

It remains merely to enumerate, as a part of our cumulative argument, some of 
the minor resemblances that link the Cratylus to its predecessors or successors, and 
make it a sort of abbreviated repertory of Platonic thoughts and classifications. In 
386 D there is a reference to the doctrine of Euthydemus: irda-i irdvra onoCoa'i elvai 
ana ical aei. In 386 D, Tr/aafets are an elSo? rmv ovtcov; cf. Thecetet., 155 E. In 387 B 
\eyei,v is TrpaTreiv, cf. Euthyd., 284 0. In 388 C ovofia dpa hihaaKaXiKov n icrriv opjavov 
Kal SiuKpiTiKov Trj<} ouCTwis, coupled with the statement, 390 B 0, that only the dialec- 
tician can use this tool, implies the imagery and doctrine of Sophist, 226-31 B, where 
the KadapcTK of dialectic and Sophistic is a branch of SiaKpcTiKrj<!. In 390 B the state- 
ment that the user is the best judge recalls Euthyd., 289 D; Rep., 601 D, and is implied 
in Phcedr. , 274 E. In 390 epcoTav Kal airoKpiveaOai eTricrTdfjievov may be compared 
with Phcedo, 75 D. In 390 D the dialectician, as iincndTri';^ suggests Euthyd., 290 C ; 
Rep., 528 B. In 392 C the view of the capacity of women is that of Rep., 455 D. 
With 394 D cf. Rep., 415 B, on the probability that good men will breed true. With 
396 C, op&cra ra dvco, cf. Rep., 509 D. In 398 A-0 the image of the golden race, and 
the identification of good men with daemons recall Repub., 415 A and 540 C. In 
398 E the rhetorician is akin to the dialectician (ipmrvriKoi epm, cf. Symp.), which 
makes against Sidgwick's view that in the earlier dialogues the Sophist is a rheto- 
rician, in the later an eristic. In 399 man is distinguished from the brute by con- 
ceptual thought, as in Phcedr., 249 B. In 400 B the conceit <TS)ix.a arj/xa repeats 
Gorgias, 493 A. In 401 B iieTeapoXdyoi, Kal aSoXeaxai, rive; is precisely in the tone 
of Phcedr., 270 A., aSoXeaxiai Kal fierecopoXoyia'; (ftvaeaxs irept,. In 401 C ovala ''Eiaria 
recalls Phcedr., 247 A. In 403, 404 characteristic doctrines of the Phcedo, Gorgias, 
and Symp. are implied concerning the naked soul, the invisible world, death, imOvfiM 
as &/7/io'?, and the yearning of the soul for pure knowledge. Cf. Gorg., 523 C; Phcedo, 
83 C D, 67 E-68 A. In 408 C the association of 7^yo<i a\r]0^'i re Kal -^jrevSTp with the 

680NATOHP, however, Archiv, Vol. XII, p. 163, thinks 682 s«pra, pp. 54, 56, 51, n. 373. 

the lack of dramatic mise en seine a mark of lateness. ''^^ Supra, p. 33, n. 218, n. 539. 

S81 See Jowett's Index, s. v. " Etymology." '''* Supra, p. 54. 

76 The Unity op Plato's Thought 

movements of the All recalls Tim., 37 B C. The quibble vi^epa, ^/lepa, 418 D, is 
repeated in Tim., 45 B. In 418 E aya6ov ISea ro Beov is explained by Eep., 336 D. 
In 419 C Xrhrr) airo rfji BiaXvaea)^ implies the doctrine of Phileb., 81 D, and Tim., 64D. 
In 422 A o-Totx«a is used for elements, as in Tim., 56 B ; Themtet, 201 E. In 423 C D 
music is "already" ij.iiitj(ti<;. In 4280 the i^airarda-dai avrov w^' avrov is virtually the 
"voluntary lie" of Bep., 382 A. In 436 D the emphasis laid on the apxv or hypothesis 
(inroicetTai) recalls Phcedo, 101 D, 107 B. 


The Euthydemus in subtlety of logical analysis, and in its attitude toward eristic, 
is akia to the Sophist and Thecetetus.^ The question. Can virtue be taught? the pro- 
treptic discourses, and the quest for the political art resume similar discussions in the 
Meno, Protagoras, Charmides, and Gorgias.''^^ To the partisans of development the 
dialogue offers a dilemma. Either this mature logic must be assigned to an early 
work, or a late work may display comic verve of style and engage in a purely dramatic, 
apparently unsuccessful, Socratic search for the political art.^' 

A systematic analysis would be superfluous after Bonitz, Grote, and Jowett. But 
the Euthydemus, like the Cratylus, is a repertory of Platonic thoughts that link it to 
"earlier" and "later" dialogues. A few of these may be enumerated: 273 C, avrov 
axnm ^or)6elv iv tow BiKaa-Tripioi<;; cf. Oorg., 509 B; 275 D, the captious question, Are 
those who learn ot a-otjiol f', ol afia6el<! ? merely illustrates the doctriae of Lysis, 218 A ; 
Symp., 203 E; Soph., 229 C, 276D ff.; do they learn a eirCaTamat fj a (iri, recalls the 
method Kara to elSevai rj firj elBevai of the Thecetetus,^'^ and the distinction between 
imaT-nM': efw and /cttjo-j?; cf. 277 C and 278 A with Themtet, 197 B; in 276 E a^vicTa 
is used as in Thecetet., 165 B; 278 B ■n-poairaii^eLV is used for eristic, as irai^eiv in 
Thecetet., 167 E; 280 E, to Be ovre icaKov ovre ayaOov; cf. Lysis, 216 D; Gorg., 
467E; 282B, ouSei/atV%/3w .... BovXeveiv .... ipaary .... TrpoOvfiovfievov a-o(f>ov 
yeve'a0ai, cf. Symp., 184 C; 284 B, \eryeiv is irpaTreiv, cf Cratyl, 387 B; 287 A, if 
there is no error, TtVo? BiBdaKoXoi rjKere, cf. Thecetet., 161 E, 178 E; 287 D, irorepov ovv 
yjrvxvv exovra voel to, voovvra. The quibble suggests the metaphysical problem of 
Parmen., 132 D, cf. A. J. P., Vol. XXII, p. 161; 289 C, the art of the user and the 
art of the maker, cf Bep., 601D, Cratyl, 390 B, 290 A, cf Gorg., 454; 290CD, cf 
Polit, 305 A, and supra, p. 62; 290 C, the mathematician subordinated to the 
dialectician, cf. Bep., 528 B; 291 B, Sxnrep ra iraiBCa to, tov^ KopvBow BimKovTU, etc., is 
the germ of the image of the aviary in the Thecetetus; 291 C, cf. Polit., 259 D ; 292 D, 
cf Charm., 167 0, Meno, 100 A, Protag., 312 D; 301 A, cf supra, n. 199; 301 B, 
cf. supra, n. 426. 

585 Sitpra, pp. 54, 58. verbreitet ist. Man sollte doch in Erw5gen Ziehen, ob 

B86 Of Idea of Good, p. 204 ; supra, a. 97. ^e^^ j^^e Euh6 und Sicherheit der Discussion einer Frage 

687 292; cf.mpra, n. 71. Bonitz, p. 125, protests against als Frage far jemand mOglich ist, Mr den sie eben nur 

the assumption that Plato is really baffled in 292 E, and ^'X'^ Problem ist und eine MOglichkeit der LOsung sich 

sensibly adds : " Ich erwahue dies nur, weil diese Art der "^'cht dargoboten hat." 

Folgerung und der Erklaning Platonischer Dialoge weit ^^^ Supra, nn. 547, 548. 

Paul Shobey 77 

The significance of the closing conversation with Crito is often missed."" Nothing, 
of course, can be inferred from the casual admission (307 A) that xpjj/toTto-TtK^ and 
jyqropiKr} are ayaSov; or from the "contradiction" of the Hepublic in the statement that 
philosophy and iroKmKr) trpa^ii are both ayadov, but -n-pm aWo sKarepa. Socrates is 
speaking to his worthy friend the business man Crito from the point of view of common- 
sense. We have also seen that the allusion to Isocrates (?) does not determine the 
date.'*" Plato is defending himself and Socrates against the criticism that such trivial 
eristic is unworthy of the attention of a man of sense. The dignified rhetorician to 
whom the criticism is attributed, like Isocrates, confounds eristic with philosophy and 
proclaims the futility of both.'" Plato replies (1) that in philosophy as in other 
pursuits the majority are bad; (2) even eristic may be a useful logical discipline. 
The second thought is implied rather than expressed. It is implied by the interven- 
tion of the Saifioviov (272 E) and by the statement that the gentlemen who in 
Prodicus's phrase °°^ hold the borderland of philosophy and politics, and who think the 
philosophers their only rivals for the first place, are badly mauled in private conversa- 
tion when they fall into the hands of eristics like Euthydemus.°'' Socrates, on the other 
hand, though ironically admitting defeat, has shown himself throughout able to do 
what is postulated of the true dialectician in the Sophist, 259 C: rots Xeyofievoi<! oldv re 
eivai Kaff eKuarov ikeyxovra eiraicoKovdelv.^ The multitude think such logical exercise 
unbecoming. But that is because, in the words of the Parmenides (136 D), ayvoova-i 
. , . . OTi dvev TavTijs t^9 Slo, irdvrcav Sie^oBov re kuI 7rXdvr}<; dBvvarov evTVXpvra tw a\r}6d 
vovv exeiv. But Socrates, regardless of personal dignity, welcomes every occasion for 
intellectual exercise: ovra rt? eptoi Seivm ivSeSvxe t^? irepl tuvtu yvnvaaCa<{ (ThecBtet, 
169 C). 


The leading ideas of these dialogues have already been studied, and it is not 
necessary to analyze them in detail.'"" We may acquiesce in the presumption that the 
Protagoras, Gorgias, and Meno are somewhat earlier in manner and style™ without 

689 Geotb, e-g., says : " In the epilogne Enthydemus is Adam (edition of Republic, index, «. v. "Isocrates"). Ob- 
cited as the representative of true dialectic and philoso- viously, barely conceivable references in Plato to an Iso- 
p]jy_ii cratean type of thought or a Qorgian style prove nothing, 

590 <i t> 72 ^°^ "*" ^lyt^'i^S be inferred from coincidence in common- 

^ ' places or in ideas that can be found in Euripides and 

Ml 305 A, ital oStoi (Dionysodoms and Enthydemus) iv Thucydides. It would be easy to "prove" by these methods 

Toit KpaTioToi! eicri rav vvv. jjjjj^i jjjg Busiria follows the Bepublic and precedes the 

B92See Joel, Der echte und der xenophontische Sohra- Sj/mposmm which contradicts it (c/. Busiria, I, -wittSymp., 

tea Vol. n, p. 634. 198 D) . Strangely enough, the very critics who force a 

, , . . ... ~. reference to the Helena upon Bepublic, 586C, are apt to 

593305D, .V Se To« cSto« X.iyo« W ..roWSo,^... Cf .^ ^^^ ^^^^^^^ ^j ^^^.^ chronology, the two almost 

Theiztet., m B or. ay .«., X070.Se, Sov... re «a. 6.fa<r9a.. The ^^^^^ ^.^^y^^^ ^j Isocrates by Plato, that in Ph<Bdr., 

rhetorician is helpless in the hands of either the philoso- ^g, a ^^pra, p. 71), and that in 0<yrgias, 463A, where Isoc. 

pher or the eristic. ^ ^^_ ^l koI ifiuxi' «i'*p'«5! t"' SoJaffriKis epyoi- t'vaji. is wittily 

B9« Cf. supra, nn. 117, 426. ''95 See Index. parodied by ^"xi' *' <rroxaori«5t koI irSpeiat. DOmmler 

696SUDHAUS, Bhein. Mua., 44, p. 52, tries to assign the calls this a "nicht einmal wOrtliche Uebereinstimmimg in 

Gorgias to the year 376 between the To Mkokles and the einem banalen Gemeinplatz." But the very point of the 

Nikokles. He is refuted by DOmklee, KleiTie Schriften, I, jest lies in the substitution of the lower word, o-roxao-rucS!, 

pp. 79 fl., who proposes other Isocratean parallels, which for the term Sofaffnic^s intentionally employed by Isocrates 

are courteously, but sensibly, minimized or rejected by to mark the superiority of his S6(a to the pretended 

78 The Unity of Plato's Thought 

admitting that there is any traceable development of doctrine.*" There is also, as 
we have seen, no evidence in the thought sufficient to date the Symposium and 
Phcedo relatively to each other or to the Republic, the Phcedrus, and the Thecetetus. 
Pfleiderer thinks the Symposium the first dialogue of Plato's "third phase," which 
includes the Philebus, Timceus, Critias, and Laws. He sees in Symp., 209-12, a 
review of Plato's previous career, with many allusions to the different "phases" 
of the Republic (p. 46). So also Dilmmler, infra, n. 619. It suffices for our pur- 
pose that all these dialogues were written after Plato had attained maturity of years, 
and presumably of thought — the Meno after 395,^" the Gorgias after Isocrates's 
Against the Sophists, the Symposium after the year 385,°°° the Phcedrus probably after 
Isocrates's Panegyricus. That the Phcedo cites the Meno is probable."'" That the 
Republic alludes to the Phcedo is possible, but not necessary;™'' and, having other 
reasons for believing the Phcedrus to be later than the Gorgias, we may assume that 
Phcedrus, 260 D, 261 A, alludes to Gorgias, 462 B, without, however, admitting the 
validity of such arguments as Siebeck's suggestion (p. 116) that Ope/xfiara yewaia 
intentionally characterizes the Xoyoi as "etwas Herangepflegtes, Ausgearbeitetes." 

But it is idle to pursue this aKiafiay^Ca further. 

The chief witness to the unity of Plato's thought is the Republic, the great work 
of his maturity and the most complete synthesis of his teaching. It is presumably 
later than most of the minor Socratic dialogues,"" but it completes rather than contra- 
dicts them, and their methods imply its results.""* It is earlier than the Laws and 
Timceus, and probably than all or most of the dialectical dialogues, but they do not 
contradict it, and they develop no important idea which it does not distinctly suggest.™' 

It is generally dated somewhere between 380 and 370, and we may say, if we 
please, that it was published when Plato was about fifty-five years of age, but any date 
between his fortieth and sixtieth year will serve as well."°° 

67rii7T^iiiT| of the metaphysicians. On the other band, though of the Pfeoedo. But Pfleidekee (p. 92) finds that "das 

the Phcedr, is in point of fact probably later, nothing can be Allegro des Symposion .... auf die schwermtltigernsten 

inferred from its agreement with Isocrates (Phcedr., 269 D; Trauerklange des vorhergehenden Sterbedialogs nunmehr 

Isoc. in Sophist^ 17) in the commonplace that eirttrr^juiij, die verklArten Harmonien einer wiedergefundenen Lebens- 

jLteXenj and <^va-ts are indispensable to the complete rhetor. stimmung folgen lasst." It's a poor argument that will not 

They are requisites of the ifcaebs avwvtffTTjs in any pursuit, work both ways 1 

as is distinctly stated in Bep., 374D E. Nor is anything to ^^^Supra, pp. 19, 40 S., 43. 699go A. 

be learned by pressing too closely the various possible 600193 A. It is, of course, just conceivable that, as WrL- 

meanings of eirwrr^nTj — knowledge of the Isocratean rules amowitz affirms (Vermes, Vol. XXXII, p. 102), the allusion 

of rhetoric, knowledge of dialectic and psychology that is to the events of the year 418. But we are still waiting 

might make rhetoric an art in Plato's opinion, knowledge for his proof that Plato commits no intentional anachron- 

of the subject-matter of the discourse. isms. 

69'Zellee says, p. 527, that the Protagoras, which ^i 73 A ; Meno, 82 ff. It is not necessary, for Plato prob- 

assumes the identity of the good and the pleasurable, ^'^'^ often illustrated ivafjvijirit by geometrical cross- 

"must" be later than Gorg., 495 ff., and aU subsequent examination in the school. 

dialogues. But c/. supra, p. 20. Hoen finds in Protag., 602iJep., 611B, oi aMat (Aoyoi) need not be the specific 

Oorg., and Phcedo the following Denkfortschritt : (1) Die proofs of immortality given in the PTkecJo. 
Lust ist das Gute. (2) Die Lust ist nicht das Gute. (3) Die 603Siebeck, however (p. 126), thinks that the Laches is 

Lust ist das BOsel In Phaedr., Symp., Phcedo he sees a the fuller discussion of courage "promised" in iSep., 430 C, 

falling away in middle life from the youthful faith in avflir 6e Trepl ovtou, ea^v ^ovKji en KaWiov SUnev. 
immortality to which age returns I Lutoslawski thinks *"* C/. supra, pp. 14, 15. 

that the discussion about the identity of the tragic and >^^ Supra, nn. 244, 375, pp. 34, 36, 42, 46, 55, 62. 

comic poet at the end of the Symposium is an apology for 606 See Zelleb (pp. 551 ff.) , who dates it in 375. The 

the comic touches in that dialogue and an announcement coincidences between the Bepublic and the Ecclesiazousae 

Paul Shoeey 79 

The relations already indicated between the Republic and other dialogues force 
extreme partisans of "development" to break it up into distinct sections which they 
assign to different periods.™' Such hypotheses are beyond the scope of serious criti- 
cism, which in the total absence of evidence can neither affirm nor deny them. It can 
only point out the fallacy of the reasoning by which they are supported. The "argu- 
ments" of Krohn, Pfleiderer, and their followers have been refuted in more than suffi- 
cient detail by Hirmer, Campbell,*' Grimmelt, and other defenders of the unity of the 
Bepublic. They may be reduced, broadly speaking, to a petitio principii and a few 
typical fallacies. The petitio principii is the assumption that the numerous connect- 
ing links and cross-references that bind together the "parts" of the Republic were 
inserted by Plato as an afterthought. The chief and fundamental fallacy is the appli- 
cation to a great and complex literary masterpiece of canons of consistency and unity 
drawn from the inner consciousness of professional philologians. The architectural 
unity of the Republic is superior to that of the Laws, the Philebus, the Phcedrus, or 
to that of the parts into which the disintegrators resolve it, many of which plainly 
could not exist by themselves. Secondary intentions, a prelude, digressions, and a 
peroration, postlude, afterpiece, or appendix may be expected in so long a work. As 
Jowett sensibly says:*"' "We may as well speak of many designs as one; nor need any- 
thing be excluded from the plan of a great work to which the mind is naturally led by 
the association of ideas and which does not interfere with the general purpose." It is 
uncritical, then, to assume a central argument and prune off everything that is not 
indispensable to its development. The argument might conceivably have started from 
the restatement of the problem by Glaucon and Adeimantus at the beginning of the 
second book. Plato might have drawn up a sketch of a reformed state, omitting all 
mention of the higher education, the rule of the philosophers, and the degenerate 
forms of government. He might have closed the work abruptly with the demonstra- 
tion of the main thesis at the end of the ninth book. Or, if he wished to add the 
myth, he might have omitted or found another place for the digression in which the 
banishment of the poets is justified on deeper grounds. But these bare possibilities 
do not raise the slightest presumption that the Republic was, in fact, pieced together 
out of detached and disjointed essays. The different topics were closely associated in 
Plato's thought. And if they were all present to his mind from the beginning, it 

of AristophaneB yield at the most a terminus post quern. the picture of the tyrant (577) "must" fall after the first 

Of HiBMEB "Entstehung nnd Komp. d. Plat. Eep.," Jaftr- Sicilian journey and before the second when Plato was on 

bUcherfar Phil Snppl., N.F., Vol. XXIU, p. 655; Adam, friendly terms with Dionysins the younger; (3) because 

The Bepublic of Plato, Vol. I, pp. 345-55. Hibmee (pp. 660 £E.) Cheist has " proved " that the eleventh epistle (circa 364) 

disposes of the attempt to date the Bepublic by the allu- is genuine, and the eleventh epistle implies the completion 

sion to Ismenias (336 A), and to Polydamas (338C), by the of the Bepublic and the beginning of the Ttmcetts. 

supposed allusion to Eudoxus (530) , and by Rbinhaedt's cot Pfleideeee, Zur Ldsung d. plat. Frage, p. 79 : " Das 

reference of 410 BC to Isocrates's Antidosis, 181, and of Znsammenwerfen ganz verschiedener Phasenin der Bep.. 

498DE to the Areopagiticus. He himself, with as little ^jg j^jj behaupte, musste nothwendig far Jeden, der sonst 

proof, thinks that 498 DE alludes to the Euagoras. He gerne Phasen und Perioden gesehen hatte, die geahnten 

dates the completion of the Bepublic circa 370 : (1) because, Greuzlinien wieder verwischen." 

after Christ, he believes that the protest against interne- ^^^ n^pMic, Vol. 11, essay III. 

cine war between Greeks (471 A-C) "must" refer to the J^i- ' .. ' 

destruction of Plattea by the Thebans in 374; (2) because ««' Vol. HI, p. vu. 

so The Unity of Plato's Thought 

would not be easy to suggest a more natural and effective order of presentation than 
that in which we now read them. 

To prove, then, that, as a matter of fact, the "parts" of the Republic were com- 
posed at different times recourse is had to two other fallacies : (1) it is assumed that 
what is not explicitly mentioned in any part is not known to the author at the time ; and 
(2) slight variations in phrasing are taken to imply serious differences of doctrine. 
The application of this method to the theory of ideas and to Plato's psychology has 
already been considered."" A few words may be added here on the second point. 
Rohde"' says that the immortality of the soul is ignored in the earliest part, II- V, 
471 C ; first appears as a paradox in X, 608 D ; and is assumed in its sublimest form 
in VI, VII. But his arguments will not bear scrutiny. "Was nach dem Tode kom- 
men moge, soUen die <f)v\aKe<i nicht beachten" (III, cap. iff.), is an unwarranted 
inference from Plato's polemic against Homeric verses that represent death as terrible 
to all men, even the good — an idea which Plato would always have repudiated. The 
sneers in 363 C D and 366 A B at future rewards are directed against low ideals — the 
fiedrjv aiwviov — or are intended to emphasize the necessity of first proving that virtue 
is desirable for its own sake. When that is done, it is nBr) ave7ri<f)6ovov (612 B) to add 
the rewards ; and there is no more inconsistency in reintroducing in a nobler form the 
premiums which the gods bestow upon virtue after death than there is in the with- 
drawal of the supposition that the just man is to be reputed unjust, and in the affirma- 
tion that in fact honesty is the best policy, though that is not the sole or the chief 
reason for practicing it."^ 

The omission of all reference to immortality in the first nine books would prove 
nothing. It is equally ignored in the first nine books of the Laws, and is first 
explicitly mentioned in XII, 959. Glaucon's dramatic surprise at Socrates's confident 
assertion of immortality proves nothing for Plato. The idea is familiar to the Gorgias 
and Meno. And even if we deny the reference of 611 B to the Phcedo, and with 
Rohde place the Phcedo after the Republic, the tenth book of the Republic knows the 
ideas, and even the rpiTO'; dvOpcoiro^, and cannot therefore be placed before the Gorgias 
by those who make use of arguments from development. In speaking of immortality 
Plato naturally tries to qualify and limit the doctrine of the tripartite soul."' He can 
only fall back upon poetical imagery and affirm his faith that in its true nature the 
(immortal part of the) soul must be one and simple. It is a waste of ingenuity to 
attempt to find a consistent chronological development in this point in the Phcedrus; 
Rep., II- V, X; Phcedo, and Timceus. It is perfectly true, as Diimmler argues,"* that 

eioSMpj-a, pp. 36,-40 S. en psycAe, pp. 588 ff. kann irgendwelche utilitaristische Begrflndung nicht 

612SIEBECK (p. 144) and DOmmlee (Vol. I, p. 248), it is mehr interessieren." Terrible logic 1 Are modem believers 

true, find fault with this too, on the ground that the i" immortality wholly indifferent to utilitarian considera- 

Socrates of the tenth book does not repeat every point of tio^is " als Zugabe " ! And had Plato no interest in the 

the hypothesis like a lawyer, and forgets the stipulation psychological proofs that the virtuous life is, even in this 

that the unjust man was to have the power, if detected, world, the most pleasurable, given in the Laws, the Phili' 

to defy punishment, or the wealth to buy off the gods. *««i and the ninth book of the iJep«6IJc? 
Dflmmler also objects that " nachdem die Perspektive auf ^^^Supra, pp. 42, 46. 

die Ewigkeit als myiuTa SSAa der Tugend bezeichnet war, 61* Vol. I, pp. 256 S. 

Paul Shoeey 81 

if the soul is really one, the definition of justice as a relation between its parts loses all 
meaning. But such "inconsistencies" are inherent in human thought, and prove 
nothing for the relative dates of Book X and Books II-V. Can any modern theo- 
logian produce definitions of the virtues that will apply to man in his earthly state 
and to the disembodied soul?"* 

Lutoslawski, while rejecting the fancies of Krohn and Pfleiderer, holds it pos- 
sible to show that the first book of the Republic falls between the Gorgias and the 
Phcedo, and that the remaining books follow the Phcedo and reveal traces of progressive 
development of doctrine. The following parallel illustrates the force of his arguments : 

P. 277: "This sharp and general formula- P. 318: ''Here"' for the first time occurs a 

tion of the law of contradiction,"" not only as formulation of the law of contradiction as a 

a law of thought as in Phcedo,^" but for the law of thought, while in the Phcedo and earlier 

first time as a law of being .... is a very books of the Republic it was a metaphysical 

important step." law." 

Lastly, a word must be said of the attempt to trace a development in Plato's 
treatment of poetry. The contradictions of those who employ this method might be 
left to cancel one another."' But the whole procedure is uncritical. Plato was always 
sensitive to poetic genius, and there was no time when he might not have praised 
Homer without conspicuous irony .*^'' But he always regarded the poet as an imitator, 
whose aim is pleasure rather than the good, whose ethical teaching must be inter- 
preted or controlled by the philosopher, and whose fine sayings are the product of 
"inspiration" rather than of knowledge. The Apology''''^ anticipates the Republic in 
the doctrine that the poets do not know whereof they speak, and the Phcedrus in the 
theory of poetic inspiration. The Gorgias, 502 BOD, deals with the moral influence 
of poetry upon the masses in the tone of the Republic and Laws; and like Republic, 
601 B, strips from the body of the poet's discourse the meretricious adornment of the 
poetic dress. The doctrine that poetry is /tti]u''?o''? is sufliciently implied in Gratylus, 
423, where the mimetic value of words is discussed, and where /xovo-ikt] is classified as 
fjiifiriai<!. The differences between the tenth and the third books of the Republic 
cannot be pressed. The third book hints that there is more to come ; ^''^ and the tenth 
book announces itself as a profounder discussion, based on psychological distinctions 
brought out in the intervening books. But it is begging the question to assume that 
they were discovered by Plato after the composition of the third book. The fact that 

615 Cf. supra, pp. 6, 7, and HiRMEE, p. 641. the Republic, and sees in it a return from the bitter mood 

.,.,„„_ (!i7inoT? 6i8Rn9'P' of the Gorpios and iSepufclic to a calmer and more generous 

state of mmd; Da ist er auch gerecht gegen andere; 
619 Lutoslawski says that Plato's scorn of poetry de- jj^^^^. ^^^ Hesiod, Lykurg und Solon sieht er unter sich, 
veloped after the Symposium, and that the tenth book of ^^^^ ^^^ ^^^^^ anderen ! " 
the Republic is therefore later than the Phcedo, -which ,„„„,,„,,- . . .« - „ - 

^ 1- -i, E>t, 7„.,- 620 p/tcedo, 95A, ovts yap av , , , , OuTjptd Petto TrOfnTn 

praises Homer without irony, and earlier than P/KEdrns , ^<t.i«uo, - . tc_ _, „„,V'' x- j, - 

,™ . ij 4.T,« 1^^ «r.+;™afa «f ouoXovoiaef OUT6 avTOt 7j/Lttf ttUTOt?; i^aWS, 776 ±/, o 6e trot^toTaTOs 

and TftecEtetMS, which take for granted the low estimate Of /*. '_ '* . . J^ ■ i iv 

„ ,V.T J.1-- 1 • X +1 r,^n™«.. rtf f>.Q »3^ttf Ttdc irotTjTuc — in both passages seriously, as the con- 

thepoet. But Natokp, thinking of other passages of the ! i. -i, 

Phcedrus, is positive that such a dialogue could not have 'ext shows. 

been written after the rejection of poetry in the Republic ; '" 23 C ; cf. the Ion. and Mem, 99 E. 

while DOmmleb (Vol. I, p. 269) places the Symposium after 622 394 D, lirm Si «at n^^eto. e-rt rourtui/. 

82 The Unity op Plato's Thought 

in emphasizing the distinction between dramatic and narrative poetry Plato carelessly 
speaks as if the former alone were imitative, proves nothing.'^' A far more important 
new point made in the tenth book is already distinctly implied in the Protagoras — 
the antithesis between the principle of measure in the soul and rj tov (f)aivofievov 
Svm/it?,"" to which poetry makes its appeal.'^^ The mood of the Symposium differs 
from that of the Gorgias and the Republic. But this does not prove either that the 
Symposium is earlier, or that Plato had been mellowed by success. A banquet at 
which Agathon was host and Aristophanes a guest was obviously not the place for a 
polemic against dramatic poetry. But even here the ironical superiority of the dialec- 
tician is maintained, and the inability of the poets to interpret or defend their art is 
revealed. °^° 


The value of Plato's life-work would be very slightly affected even if it were true 
that in the weakness of extreme old age the noble light of his philosophy did " go out 
in a fog of mystical Pythagoreanism." It is not in the least true, however, and the 
prevalence of the notion is due mainly (1) to the uncritical acceptance of the tradition 
concerning Plato's "latest" doctrine of ideas and numbers; and (2) to the disparaging 
estimate of the Laws expressed by those who care only for dramatic charm of style, or 
by radicals like Grote, who are offended by the " bigotry" of a few passages. A word 
must be said on each of these points. 

1. Aristotle's account of Plato's later identification of ideas and numbers has 
been generally accepted since Trendelenburg's dissertation on the subject.'^" Zeller 
rightly points out that the doctrine is not found in the extant writings, but adds that 
for Plato numbers are entities intermediate between ideas and things of sense. In 
my discussion of the subject ^'^ I tried to establish two points : first, that we need not 
accept the testimony of Aristotle, who often misunderstood Plato, and was himself 
not clear as to the relation of mathematical and other ideas; second, that the doctrine 
of numbers as intermediate entities is not to be found in Plato, but that the passages 
which misled Zeller may well have been the chief source of the whole tradition about 
ideas and numbers. The first point is a matter of opinion. I did not deny the testi- 
mony of Aristotle, and no one who chooses to accept it can be refuted. The relation 
of ideas to numbers was doubtless much debated by the scholastics of the Academy. 
Aristotle's reports of the intolerable logomachy do not make it clear just how much 
of this nonsense he attributed to Plato. But I do not intend to enter upon the inter- 
pretation of the eleventh and twelfth books of the Metaphysics. No reader would 

623 393 C, 394 D. 62*Pro«aff., 356D. man is "inspired" by the tragic muse, another by the 

625 Rep., 602, 603. comic. If poetry were a matter of science, the poet could 

626 201 B, KH-Svi/evio, S> Sai/(pare9, oiSiv eiSivat mv Tore eiiroi-. '^^e both forms, even as the scientific interpreter of poetry 
Kai fiT|v KaAis ye elTTes, <t>avai., S> "Ayieui/. Cf. also 223 D, where would not, like the " inspired " Ion, be limited to Homer. 
Socrates compels Agathon and Aristophanes to admit toC This we may plausibly conjecture to be the meaning. But 

avTOv avSpo^ elvai, KtafitoSiav Kai TpaytoSiav iiriirrairOai, iroicl*'. *t IS only conjecture. 

This is thought to contradict Bepub.^ 395 A, but the contra- ^^"^ Plat, de id. et numeris doctritia, 1828. 

diction is removed by pressing rix<'ri in what follows. One 628 Oe Plat, id, doctrina, pp. 31 H. 

Paul Shoeey 83 

follow me, and no results could be won. If Aristotle's testimony be accepted, there is 
an end of controversy. Plato taught in his lectures the doctrine of ideas and numbers. 

But the second point is not so elusive. It is possible to test the argument that 
the extant writings do not recognize an intermediate class of mathematical numbers, 
and yet might easily suggest the notion to mechanical-minded students. Now Zeller 
in his fourth edition confounds the two questions. He gives the impression that he 
is answering me by a Quellenbelege from Aristotle and Philoponos. He wholly 
ignores my interpretation of a number of specific Platonic passages, which he appar- 
ently takes for the mere misunderstandings and blunders of a beginner.'^^ I have no 
hope of convincing Zeller, nor do I wish to force myself into a polemic with the 
honored master of all who study Greek philosophy. But, as Mr. J. Adam, a scholar 
whose scrupulous candor makes it a pleasure to argue with him, has expressed surprise 
in his edition of the Republic that I still adhere to my opinion in spite of the mass of 
evidence, I will endeavor tb state my meaning more plainly. 

The theory of ideas, the hypostatisiation of all concepts, once granted, numbers 
do not differ from other ideas. The phrase, irepl airrSiv rmv apiOfJ^wp (Rep., 525 D), 
denotes ideal numbers or the ideas of numbers, and opara fj aina aw/jLara exovrai 
apidfiov<} are numbered things, things of sense participant in number.**' That is all 
there is of it, and there is no extant Platonic passage that this interpretation will not 
fit. For educational purposes it is true that mathematical science holds an inter- 
mediate place between dialectic and the perceptions of sense. Mathematical 
abstractions [r/ irepl to ev iid6rj<ri<;, Rep., 525 A) are the best propsedeutic to abstract 
reasoning generally. But there is no distinction of kind between them and other 
abstractions, ff«Xi7poi' /[toXa/coV (Rep., 524 A ff.). Mathematical science as Bidvoia is 
midway between the pure vow of dialectic and the Bo^a of -sense. But that is because 
of its method — the reliance on diagrams (images) and hypotheses. In themselves its 
objects are explicitly stated to be pure voijTa."^^ The "mathematical" numbers then 
are plainly the abstract, ideal numbers of the philosopher. The numbers of the 
vulgar are concrete numbered things. There is no trace of a third kind of number."^'' 
Those who have not yet learned to apprehend abstractions mockingly ask the mathe- 

«29 It may be permissible to add that he seems to have the v<ntra being divided into two classes by re-ztai. The 

read other parts of the dissertation with more attention, sentence still stands, and I am quite willing to leave the 

since, to mention only two cases, he adds on p. 745 a refer- question of Flilchtigkeit to any competent scholar, e. g., to 

ence & propoB of the Tpiroi ai-Spujro! to Bepublic, 596, 597, M. Bodier, who translates " les intelligibles, anssi bien les 

and Tim., 31 A, with the interpretation of their significance concepts abstr-.its (on mathfimatiqnes) que (ceox qui ont 

given on p. 30 ; and he omits from p. 547 of the third edition pour objet) les qnaiitfis, etc." 

a sentence criticised on p. 49 of the dissertation. Another 530 ^^^m translates avTai" tok opiS^ii^, " numbers them- 

slight but significant point may be mentioned. Aristotle selves," which is quite right. My point is that " numbers 

himself makes a not wholly clear distinction between themselves " are proved by the context and by Philebus, 

mathematical ideas {rH iv i^aipeacL Aeyd/iera, almost tech- gg j,^ j^ ^^ y^^j numbers. For Adam's further argument 

nical) and other ideas. In illustration of this I objected ^^ infra, p. 84. 
to Zeller's interpretation of De ^jt., 432a2, ei- TOW eiaeo-i To« „, „ ,,,„t^ . . ... •» 

abstrakten Begrifle ") ital oo-a Tuv aicrSi)™!' efew KOI ira»r). fitrpuv ou7,/», .^ " i> / 71- V 1 

My objection was that both grammar and Aristotelian °''™'' '"'■<' "CXI!- 
usage showed that otra rdv oltrB-riTiov, etc., are also abstralc- ^^^ Fhileb., 56 BE. 

te Begriffe (in the German or English sense of the words). 

84 The Unity of Plato's Thought 

maticians (Rep., 5 26 A), Trept iroloov apiOfimv SiaXe^eade; and the answer is, irepl tovtcov 
wv BiavorjOfjvai fiovov iyx^P^h coupled with an exposition that recalls the Parmenides 
of the pure idea of unity .''^ Simple as all this appears, it might easily be misunder- 
stood by the pupils of the Academy. Mathematics was intermediate from an 
educational point of view. In cosmogony numbers and geometrical forms are the 
mediators between chaos and the general idea of harmony and measure."^* The 
expression, numbers (arithmetic), of the vulgar and numbers of the philosopher would 
lead a perverse ingenuity to ask of the mathematicians, in the words of the Republic, 
Trepl iromv apiOfiSiv SiaXejecrOe ; Plato's use of "dyad" and "triad" as convenient 
synomyms for the pure idea of two and three would be mistakenly supposed to imply 
a distinction."^^ The innocent question {Rep., 524 C), rt ovv ttot' eVrt ro fii'ya av koI to 
(TfiiKpov,^^ would suggest that it was a terminus technicus for some mysterious ultimate 
philosophical principle, and set students upon hunting it and its supposed synomyms 
through the dialogues, and, inasmuch as /leya -f- a/jLCKpov indubitably = 2, it might 
well be identified with the indeterminate dyad and its supposed equivalents, or any 
other "principle" posited in antithesis to the one."" The folly once set a-going, there 
are no limits to its plausible developments. All the unanswerable questions as to the 
relation of ideas to things may assume special forms for special classes of ideas. Plato 
himself shows this for ideas of relative terms in a much misunderstood passage of the 
Parmenides.^" The problem of the relations of numbered things, of the supposed 
mathematical numbers, and of ideal numbers, offered a rich feast for the quibblers and 
the oyfrifiadeh of the Academy. "Before and after" is essential to number, but there 
is no "before and after" in the ideas. Multiplicity is inherent in number, but the 
"idea" even of a million must be one. Other ideas may be imperfectly copied by 
things, but is not the number five entirely present in five things ? Echoes of this 
pitiful scholasticism are preserved for us in the metaphysics of Aristotle. But what 
possible reason can there be for attributing it to Plato? Adam himself (Vol. II, p. 
160) repeats the disconsolate question: irepl ttoicov apiOyiSiv ^ia\e<yea6e ev oh to ev olov 
vnei'S a^iovTe eanv, Icrov Te eKacrTOv ttclv Trainl Kal ouSe afUKpov Siacjiepov ; and asks: 
"Are we then to suppose that there are many ideas of 'one' ?" The answer is: "Yes, 
precisely, to the extent that there are many ideas of anything." We have already 
seen (Rep., 476A) that every idea is per se one, and yet, not merely as reflected in 
phenomena, but t^ aWi^Xcop Koivavia appears many. The contradiction is inherent in 
the theory of ideas. As against the multiplicity of phenomena, we insist on the indi- 
visible unity of the idea. But when we find the idea involved with other ideas in a 
number of instances, we are forced to use the plural. Plato does not, however, here 

633 0/. Idea of Good, p. 222; Phileb., 56 E, ei i^v imviSa "Again, great and small, swift and slow are allowed to 
liovaSot enoo-TTj! Tail' /xvfiiojr (iTjSefiiiiv oAAijj' iKKfii Sia(/iEpovi7dv Ti! exist nowhere without the mind, being entirely relative, 
eii«i. and changing as the frame or position of the organs of 

634 Tim., 53 B ff. ; Phileb., 66 A. sense varies." 

635 Phcedo, 101 C ; Parmen., 149 C j Phcedo, 104. 637 De Plat, id., p. 37. 

636 Plato is using the terms precisely as Berkeley does 638 133 C ff. ; cf. A. J. P. Vol. IX p. 288. 
when he says (Principles of Human Knowledge, XI): 

Paul Shoeey 85 

in terms pluralize the "one." He says: Of what numbers do you speak in which the 
one, i. e., the idea of one, present in each as a constituent and essential part of the 
more complex idea, etc. ? Of course, this implies a multiplicity of units in each num- 
ber, and still more in all; but only as any idea is multiplied when it appears in a 
number of others. The multiplication of the idea tt) tS)v awnaTcov KoivwvCa is more 
easily evaded than that t^ dXKrjkav Koivavia, because in the first case we may use the 
imagery of pattern and copy, while, in the second case, the idea is an essential con- 
stituent part of that into which it enters. In the special case of numbers, the paradox 
is still more glaring. But Plato is not one to be frightened from the path of philo- 
sophical consistency by a paradox which he rightly regarded as largely verbal. In 
the Parmenides he amuses himself by showing that the idea of " one " itself apprehended 
TTj hiavoCa fwvov Kaff" avTo breaks up into many."'' This does not make it the less 
necessary for the mathematician to apprehend the pxire absolute idea of unity and 
restore it as fast as it is disintegrated by analysis or the senses."" 

2. Despite many passages of stately and impressive eloquence, the Laws will 
remain the type of "frigidity" for those who, like Lucian, read Plato mainly for the 
dramatic vivacity of the Phcedrus or the artistic beauty of the Symposium. Our 
purpose is not to deny the altered mood and style that mark the masterpiece of Plato's 
old age, but merely to protest against the notion that it may be safely neglected by 
the serious student, or that it presents a doctrine essentially different from that of the 

If Plato was not to rewrite the Republic, it was almost inevitable that his political 
studies should assume the form of a project of detailed legislation for a possible Greek 
city. But even here, while recognizing that many of his theoretic postulates will have 
to be mitigated in practice,"' he holds fast in principle to the ideals of the earlier 
work.''^^ A harmony of the Laws and Republic, however, though not a difficult task, 
would demand more space than can be given to it here. We need not delay to examine 
the contribution of the Laws to our knowledge of Greek institutions, or the very con- 
siderable influence which it exercised upon the speculations of Aristotle and later 
Greek thinkers. One service which it renders to students of the dialogues we have 
already often noted. 

As the years wore on, Plato naturally grew weary of Socratic irony, of the game 
of question and answer, of the dramatic illustration or the polemical analysis of eristic. 
Even in the earlier dialogues he sometimes evades or contemptuously explains away 
an equivocation which elsewhere he dramatically portrays or elaborately refutes.*" In 
the Laws this is his habitual mood,"" and in consequence the Laws may often be 
quoted for the true Platonic solution of problems which Socratic irony or dramatic art 
seems to leave unsolved in the earlier dialogues.'" 

While acknowledging this change of mood, we must be on our guard against the 

639U3A,144E. 6«Bep.,525E; sjtpra, n. 647. 6«746. 6" 627 B, 627 D, 644 A, 864 B. 

642 739 c ff., 807 B. 6i3Eep., 436 C D E, 437 A, 454 A; e*:- Supra, pp. 13, 19, nn. 70, 71, 293. 

Cratyl, 431 A ; Symp., 187 A ; Euthyd.. 277 E. 

86 The Unity of Plato's Thought 

exaggeration of its significance by Grote, Mill, and Gomperz. Grote had little appre- 
ciation of Plato's substantive thought at any stage. He cared only for the dramatic 
illustration of the elenchus. This, which for the author was a means to an end, was 
for him the real Plato. The exposition of positive doctrine he treats as the work of a 
totally different person — a dogmatic Plato who has "ceased to be leader of the oppo- 
sition and passed over to the ministerial benches." This view, which appears even in 
Grote's treatment of the Gorgias and Thecetetus, is still more prominent in his criticism 
of the Republic. In the case of the Laws this feeling is intensified by the deep 
repugnance aroused in Grote's mind by Plato's whimsical provisions for the conversion 
or punishment of those who denied the truths of natural religion or traded upon the 
superstitions of the vulgar."** He cannot speak of the Laws without alluding to that 
imfortunate page; and the vision which he conjured up of the aged Plato as the 
Torquemada of a Pythagorean mysticism makes him totally blind to the real signifi- 
cance of what in wealth of content is Plato's greatest work. This view was accepted 
by Mill from Grote, and by Gomperz from Mill, and it leads them both to misappre- 
hend the true relation of the Laws to the Republic. Mill says: "Tn his second 
imaginary commonwealth, that of the Leges, it [dialectic] is no longer mentioned; it 
forms no part of the education either of the rulers or of the ruled.""' Similarly 
Gomperz :*'* "Plato in his old age grew averse from dialectic. In the Laws, the last 
product of his pen, he actually turned his back upon it and filled its vacant place at 
the head of the curriculum of education with mathematics and astronomy.""' These 
statements, even if we concede that they are true in a sense to the letter, convey a 
totally false impression, as a slight study of the last pages of the twelfth book of the 
Laws will show. Plato does not care to rewrite the sixth and seventh books of the 
Republic. But he defines as clearly as in the earlier work the necessity and function 
of dialectic and the higher education in the state. Even in the first book we are fore- 
warned that to complete the organization of the state the founder must set over it 
^vXaKa'i .... TOiK fiev Bia <^povrj(Tea><; tov? Se St' akri6ov<s h6^r]<! t'oWa?.*^" In the twelfth 
book we are introduced to these guardians who are to possess knowledge and not merely 
right opinion. They compose a nocturnal council which is to be the anchor of the 
state."^' Recurring to the imagery and the manner of the early dialogues,"^^ Plato tells 
us that as the pilot, the physician, the general represent intelligence {vow') applied to 
the definite ends of their respective arts, so this highest council is the head, the soul, 
the mind of the state, possessing knowledge of the political o-«o7ro? or true end of rule.'*' 

646 908-10. 647 x)ms. and KiscMSs., Vol. IV, p. 289. mentioning any other element of the higher education. 

648 Greek Thinkers, Translation, p. 466. "^^^ possessors of ^povTjo-w will surely be able /car' elSij ^tj- 

cjon, 1-1 a ir7„.,„„ occ ortc Teix (630 E) and wiU practice the dialectical methods of the 

649 To like effect Zellee, pp. 955, 956. „ ii.c' iw ui^ >, j t, ,-^- r, > i.i ^ 

recent Sopftist, Phtlebus, Pohticus. Zelles's attempt 

650 632 C. The parallelism with the Republic is obvious. t^ distinguish between ((.pdi/i),r« and the voSt of the Republic 
There, too {412 A, 497 C D) , there is a similar anticipation of jg ^ false point. <^i>ovr,aK is used in PAcedo, 69 B. 

the need of guardians who know as distinguished from the g^j ^^ „ 

assistants. In Laws, 818 A, there is another anticipation 

of the higher education. Mathematics only is mentioned «»2Proto0., 311 B; Gorff., 447, 448, 449 E ; Sitaj/d., 291 C ; 

because Plato is explaining that it is not needful for the Bep., 333. 

multitude to study it profoundly. There is no occasion for 653961,962. 

Paul Shorey 87 

No state can prosper or be saved unless such knowledge resides in some part of it as a 
<f>v\aKTripiov.^ The beginning of such knowledge is to fir] -rrXavaaOai tt/so? TroWa 
a-Toxa^ofievov aW ek ev ^ejrovra, etc/°^ Now to, t&v iroXeav vo/xifia aim at many 
things — wealth, power, and tov ikevdepop S^ ^lov!'^ Our aim is virtue. But virtue is 
both four and one. The intelligent physician can define his one aim. Must not the 
intelligent ruler be able to define his? It is easy to show how the four virtues are 
many. To exhibit their unity is harder.^" A man who amounts to anything must 
know, not only the names, but the \6''jo<s of things. And the true guardians, teachers, 
and rulers of a state must not merely rebuke vice and inculcate virtue, but they must 
be able to teach rjv Bvvafuv e%et.^' The state may be likened to the body, the younger 
guardians to the senses in the head, the elders to the brain.*'^" They cannot all be 
educated alike. Therefore Ireov apa eVi riva aKpi^ea-repav TraiSeiav ri}? efiirpoaOev.^ 
This is the education already glanced at in our phrases about the unity of purpose. 
The essence of the more accurate method is our old acquaintance to tt/jo? p,lav ISe'av 
€K tS)v ttoW&v Kal avo/jioimv BvvaTov elvai ^XeTreiv.^^ The guardian must be able to do 
what Meno could not do — ISetv TrpaiTOv, o ri irore Sia TrdvTwv tS)v reTTaprnv ravrov 
Tvy^^avei.^^ And similarly Trepl koKov re Kal ayaOov and irdmwv tmv airovSaiav, they 
must not only know in what sense each is one and many, but they must be able to 
expound their knowledge— t^v evhei^iv Ta> Xo^w ivSeiKvvaOai.^^ The thing being so 
clearly indicated, it would be pitiful quibbling to object that the word BiaXeKTiKr) 
does not happen to occur here. Its omission is possibly due to the fact that the 
Athenian throughout the Laws talks down to the level of his unsophisticated Spartan 
and Cretan interlocutors. Mathematics and astronomy, then, are not substituted for 
dialectic, but are added for a special reason among the airovSaia which the guardians 
must understand with real knowledge. The multitude may follow tradition. The 
guardians must be able to demonstrate the truths of natural religion, as we have done.*" 
Astronomy, the study of the ordered movements of the heavens, is a great aid to this. 
With astronomy is involved the necessary mathematics, which also in their relation to 
music and the arts are of use to him who is to shape the characters and laws of men.**^ 
He who cannot learn these things can never be a ruler, though he may be an assistant. 
In the last two pages of the Laws Plato evades giving a detailed account of the 
curriculum of the higher education thus indicated — perhaps he was weary, perhaps 
he did not care to repeat the Republic.'^ In any case, there is no justification for the 
statement that the Laws ignore the higher education of the rulers or substitute in it 
mathematics and astronomy for dialectic. On the contrary, the unity of Plato's 

654 962 C J cf. Bep., 424 C. **■ Cf, PTiwdr., 265 D ; and with rauTTj! ouk iari. na^tirripa 

655962D. (»e9oSo!, cf. Phileb., 16B; Phoedr., 266 B. 

656C/. ilep., 563A, ivaSij iXevflepo! B. 662965 D. Meno, 74 A, r'riv Si ilUv, i Sii izivTuii' TOVTiav 

65' Of. Phileb., 18 E, nm eo-Tix Ix ical TToAAi airHv tKirepov. iariv, oil ivvilt-aOa. avtvptlv. 

658 964 C ; cf. Rep., 366 E, rj avrov Svvi.iJ.ii if rf toC cxoi-tos 663 966 B. 
ifmxB ei-oi/. *** In Book X. 

659 964 E ; c/. Tim. , 69, 70. «65 967 E. 
•60965A. 666968D. 


The Unity of Plato's Thought 

thought is strikingly illustrated by his return in the pages just analyzed to some of 
the favorite ideas of the Republic and earlier dialogues. °" 

It is not necessary to prolong this study. The Timceus, so far as it affects our 
argument, has already been considered.""' The TimcBUS as a whole I have studied 

The object of this discussion and the expression "unity of Plato's thought" may 
easily be misunderstood. I may therefore be permitted, in conclusion, to repeat that 
I have not meant to sophisticate away the obvious and inevitable variations in Plato's 
moods, and minor beliefs from youth to old age. Nor in the study of such develop- 
ment would I reject the aid of a sober and critical method of style statistics."" My 
thesis is simply that Plato on the whole belongs rather to the type of thinkers whose 
philosophy is fixed in early maturity (Schopenhauer, Herbert Spencer), rather than to 
the class of those who receive a new revelation every decade (Schelling). And I have 
tried to show that the method which proceeds on the contrary assumption leads to mis- 
interpretation of his writings. The illustrations given are merely typical. There has 
been no attempt to catalogue exhaustively the opinions of contemporary Platonists. 
The polemic is, I trust, not discourteous, and is, I am sure, not intentionally disloyal. 
In any case, it turns generally on the meaning or relevancy of specific passages and can 
easily be tested. Some excuse for its prominence may be found in Mill's statement 
that "there are few, if any, ancient authors concerning whose mind and purpose so 
many demonstrably false opinions are current, as concerning Plato." 

667GOMPERZ supports his view of the anti-dialectical 
tendency of Plato's mind in the Laws by the hostility of 
the Sophist to every kind of antilogy. But surely eristic is 
one thing and dialectic another. The true Socratic elen- 
chus is described and the diificulty of distinguishing it 
from eristic indicated in a locus classicus in the Sophist 
(230BfE.); and hoih the Sophist and the PoUticus employ 
the keenest dialectic in order to meet and defeat eristic on 
its own ground iSoph., 259 CD). In the Philebus, which 

Goraperz thinks late, dialectic is still the highest science 
of truth {Phileb., 58). But Plato had other interests than 
dialectic, and it is unreasonable to expect him to fill the 
Laws and Timceus with repetitions of what had been said 
once for all in the Sophist, PoUticus, and Philebus, 

668 Supra, p. 37. 669 a. J. P., Vol. IX, pp. 395 £f 

6'OAs, e. g., that of Eittee, "Die Sprachstatistik in 
Anwendung auf Platon und Goethe," N&ue Jahrbiicher 
etc., 1903. 


AfilivriiTK, 32, 43, 44, 19, n. 109. 

AvdptajTOq fL€Tpov, 67. 

Cardinal virtues, 12. 

Charmides, 15. 

Copula, ambiguity of, 54. 

Courage, 11, 15, n. 43. 

Cratylus, 75 £E., 54, 56. 

Definition, 13, 16, 66, n. 86; 
by dichotomy, 50 S. 

Dialectic, 74, 86 ffi : negative 
goes too far, 17 S. 

Dichotomy, 50 S. 

Eristic, 13, 14, 19 n. 108, 50, 77, 

Ethics, 9 ff. 

Euthydemus, 76 ff., 54. 

Euthyphro, 12, 31. 

Fallacies of Plato, inten- 
tional, 4, 6, 20, 23 n. 137, 54, 
57 n. 426, n. 32, n. 42, n. 70, 
n. 106, n. 528, n. 548. 

Freedom of the will, 9. 

Generalizations, n. 500. 

Good, idea of, 16 ff., 74. 

Gorgias, 22, 31, 32, 25 n. 167, 77. 

Government, forms of, 62. 

Happiness and virtue, 25 ff. 

Hedonism, 20. 

Hedonistic calculus, 23. 

Heraclitus, 28, 68. 

Hippias Minor, 10 n. 38. 

Ideas, theory of, 27 ff. ; diffi- 
culties of, 36, 52, 84; not 
concepts merely, 29, 30, 38, 
39; not thoughts of God, 
30, 38, 65j n. 512 ; origin of, 
32; in late" dialogues, 
37 ff. ; and numbers, 83 ff. 

Isocrates, 71, 72, 77. 

Laches, 15, n. 603. 

Laws, 85 ff. 

Lie, voluntary and involun- 
tary. 10 n. 38. 

Love (epu/?), 19 ff. 

Lysis, 18, 19. 

Materialism, 68, 69, n. 283. 

Matter or space, not firi 6v, 
38, n. 502, n. 503. 

Merw, 32, 33, 77. 

Mi) Of, fallacy of, 53 ff. 

Method (icaT eifiij, etc.) 51. 
Minor dialogues, dramatic, 

13, 15. 
Negative dialectic, goes too 

far, 17 ff. 

'OvofJ-ara and p^/xara, 56. 

Parmenides, 57 ff., 34, 36, 37. 

ndvTO. pel, 68. 

Phcedo, 35, 41, 77. 
Phoedrus, 71 ff., 19. 
Philebus, 17 n. 93, 22, 23, 43, 

46, 63 ff. 
Pleasure, 22; negativity of, 

23, 24 ; in mind, not body, 

Poetry, Plato's attitude- 

toward, 81 ff. 
Political art, 17, 62. 
PoUticus, 60 ff., 44. 
Protagoras, 12 n. 48, 20, 21, 

Psychology, 40 ff., 66 ff. 

(Thecetetus) . 
Psychological terminology, 

47 ff. 

Republic, 78 ff., 14, 26, 35 n. 
238, 41, 42, 55, 51, n. 375. 

Socratic ignorance, 6; in 
minor dialogues drama- 
tic, 15. 

Socratic paradoxes, 9. 

Soul, immortality of, 40 ff., 

Soul, tripartite, 42. 

Sophist, 50 ff. 

Sprach-Statistik, 1, 5, 7 n. 

SufaiTioi/, n. 461. 

2u0po(riivTj, 15. 

Symposium, 19, 77. 

Temperaments, the two, 12 
n. 59, 13 n. 70, 62 n. 481. 

Thecetetus, 66 ff., 33, 55. 

Timceus^ 37, 88. 

Utilitarianism, 20 ff. 

Vice, involuntary, 9. 

Virtue, is knowledge, 10; 
unity of, 10; coincides 
with happiness, *25 ff.